/normxcorr/trunk : revision 1

1

<html xmlns:v="urn:schemas-microsoft-com:vml"

2

xmlns:o="urn:schemas-microsoft-com:office:office"

3

xmlns:w="urn:schemas-microsoft-com:office:word"

4

xmlns:st1="urn:schemas-microsoft-com:office:smarttags"

5

xmlns="http://www.w3.org/TR/REC-html40">

6

7

<head>

8

9

10

11

12

13

14

<!--[if !mso]>

15

<style>

16

v\:* {behavior:url(#default#VML);}

17

o\:* {behavior:url(#default#VML);}

18

w\:* {behavior:url(#default#VML);}

19

.shape {behavior:url(#default#VML);}

20

</style>

21

<![endif]-->

22

<title>Digital Image Correlation / Tracking with Matlab</title>

23

<o:SmartTagType namespaceuri="urn:schemas-microsoft-com:office:smarttags"

24

name="place"/>

25

<o:SmartTagType namespaceuri="urn:schemas-microsoft-com:office:smarttags"

26

name="PlaceName"/>

27

<o:SmartTagType namespaceuri="urn:schemas-microsoft-com:office:smarttags"

28

name="PlaceType"/>

29

<o:SmartTagType namespaceuri="urn:schemas-microsoft-com:office:smarttags"

30

name="City"/>

31

<o:SmartTagType namespaceuri="urn:schemas-microsoft-com:office:smarttags"

32

name="State"/>

33

<o:SmartTagType namespaceuri="urn:schemas-microsoft-com:office:smarttags"

34

name="country-region"/>

35

<!--[if gte mso 9]><xml>

36

<o:DocumentProperties>

37

<o:Author>chris</o:Author>

38

<o:LastAuthor>Chris Eberl</o:LastAuthor>

39

<o:Revision>2</o:Revision>

40

<o:TotalTime>143</o:TotalTime>

41

<o:LastPrinted>2006-09-21T02:40:00Z</o:LastPrinted>

42

<o:Created>2006-12-29T18:40:00Z</o:Created>

43

<o:LastSaved>2006-12-29T18:40:00Z</o:LastSaved>

44

<o:Pages>1</o:Pages>

45

<o:Words>5883</o:Words>

46

<o:Characters>33535</o:Characters>

47

<o:Company>JHU</o:Company>

48

<o:Lines>279</o:Lines>

49

<o:Paragraphs>78</o:Paragraphs>

50

<o:CharactersWithSpaces>39340</o:CharactersWithSpaces>

51

<o:Version>11.5606</o:Version>

52

</o:DocumentProperties>

53

</xml><![endif]--><!--[if gte mso 9]><xml>

54

<w:WordDocument>

55

<w:HyphenationZone>21</w:HyphenationZone>

56

<w:ValidateAgainstSchemas/>

57

<w:SaveIfXMLInvalid>false</w:SaveIfXMLInvalid>

58

<w:IgnoreMixedContent>false</w:IgnoreMixedContent>

59

<w:AlwaysShowPlaceholderText>false</w:AlwaysShowPlaceholderText>

60

<w:Compatibility>

61

<w:BreakWrappedTables/>

62

<w:SnapToGridInCell/>

63

<w:WrapTextWithPunct/>

64

<w:UseAsianBreakRules/>

65

<w:UseWord2002TableStyleRules/>

66

</w:Compatibility>

67

<w:BrowserLevel>MicrosoftInternetExplorer4</w:BrowserLevel>

68

</w:WordDocument>

69

</xml><![endif]--><!--[if gte mso 9]><xml>

70

<w:LatentStyles DefLockedState="false" LatentStyleCount="156">

71

</w:LatentStyles>

72

</xml><![endif]--><!--[if !mso]><object

73

classid="clsid:38481807-CA0E-42D2-BF39-B33AF135CC4D" id=ieooui></object>

74

<style>

75

st1\:*{behavior:url(#ieooui) }

76

</style>

77

<![endif]-->

78

<style>

79

<!--

80

/* Font Definitions */

81

@font-face

82

{font-family:Wingdings;

83

panose-1:5 0 0 0 0 0 0 0 0 0;

84

mso-font-charset:2;

85

mso-generic-font-family:auto;

86

mso-font-pitch:variable;

87

mso-font-signature:0 268435456 0 0 -2147483648 0;}

88

/* Style Definitions */

89

p.MsoNormal, li.MsoNormal, div.MsoNormal

90

{mso-style-parent:"";

91

margin:0in;

92

margin-bottom:.0001pt;

93

mso-pagination:widow-orphan;

94

font-size:12.0pt;

95

font-family:"Times New Roman";

96

mso-fareast-font-family:"Times New Roman";

97

mso-ansi-language:DE;

98

mso-fareast-language:DE;}

99

p.MsoFooter, li.MsoFooter, div.MsoFooter

100

{margin:0in;

101

margin-bottom:.0001pt;

102

mso-pagination:widow-orphan;

103

tab-stops:center 207.65pt right 415.3pt;

104

font-size:12.0pt;

105

font-family:"Times New Roman";

106

mso-fareast-font-family:"Times New Roman";

107

mso-ansi-language:DE;

108

mso-fareast-language:DE;}

109

a:link, span.MsoHyperlink

110

{color:blue;

111

text-decoration:underline;

112

text-underline:single;}

113

a:visited, span.MsoHyperlinkFollowed

114

{color:purple;

115

text-decoration:underline;

116

text-underline:single;}

117

/* Page Definitions */

118

@page

119

{mso-footnote-separator:url("Correlation_Guide_2_files/header.htm") fs;

120

mso-footnote-continuation-separator:url("Correlation_Guide_2_files/header.htm") fcs;

121

mso-endnote-separator:url("Correlation_Guide_2_files/header.htm") es;

122

mso-endnote-continuation-separator:url("Correlation_Guide_2_files/header.htm") ecs;}

123

@page Section1

124

{size:595.3pt 841.9pt;

125

margin:1.0in 1.25in 1.0in 1.25in;

126

mso-header-margin:.5in;

127

mso-footer-margin:.5in;

128

mso-title-page:yes;

129

mso-even-footer:url("Correlation_Guide_2_files/header.htm") ef1;

130

mso-footer:url("Correlation_Guide_2_files/header.htm") f1;

131

mso-paper-source:0;}

132

div.Section1

133

{page:Section1;}

134

/* List Definitions */

135

@list l0

136

{mso-list-id:115176168;

137

mso-list-type:hybrid;

138

mso-list-template-ids:1885226484 67567631 67567641 67567643 67567631 67567641 67567643 67567631 67567641 67567643;}

139

@list l0:level1

140

{mso-level-tab-stop:.5in;

141

mso-level-number-position:left;

142

text-indent:-.25in;}

143

@list l1

144

{mso-list-id:725103925;

145

mso-list-type:hybrid;

146

mso-list-template-ids:-1430781362 894173392 67567619 67567621 67567617 67567619 67567621 67567617 67567619 67567621;}

147

@list l1:level1

148

{mso-level-start-at:0;

149

mso-level-number-format:bullet;

150

mso-level-text:-;

151

mso-level-tab-stop:.5in;

152

mso-level-number-position:left;

153

text-indent:-.25in;

154

font-family:"Times New Roman";

155

mso-fareast-font-family:"Times New Roman";}

156

@list l2

157

{mso-list-id:1467965214;

158

mso-list-type:hybrid;

159

mso-list-template-ids:-1054988678 -152814394 67567619 67567621 67567617 67567619 67567621 67567617 67567619 67567621;}

160

@list l2:level1

161

{mso-level-start-at:0;

162

mso-level-number-format:bullet;

163

mso-level-text:-;

164

mso-level-tab-stop:.5in;

165

mso-level-number-position:left;

166

text-indent:-.25in;

167

font-family:"Times New Roman";

168

mso-fareast-font-family:"Times New Roman";}

169

@list l2:level2

170

{mso-level-number-format:bullet;

171

mso-level-text:o;

172

mso-level-tab-stop:1.0in;

173

mso-level-number-position:left;

174

text-indent:-.25in;

175

font-family:"Courier New";}

176

@list l2:level3

177

{mso-level-number-format:bullet;

178

mso-level-text:\F0A7;

179

mso-level-tab-stop:1.5in;

180

mso-level-number-position:left;

181

text-indent:-.25in;

182

font-family:Wingdings;}

183

@list l2:level4

184

{mso-level-number-format:bullet;

185

mso-level-text:\F0B7;

186

mso-level-tab-stop:2.0in;

187

mso-level-number-position:left;

188

text-indent:-.25in;

189

font-family:Symbol;}

190

@list l3

191

{mso-list-id:1996179431;

192

mso-list-type:hybrid;

193

mso-list-template-ids:1916686938 67567631 67567641 67567643 67567631 67567641 67567643 67567631 67567641 67567643;}

194

@list l3:level1

195

{mso-level-start-at:2;

196

mso-level-tab-stop:.5in;

197

mso-level-number-position:left;

198

text-indent:-.25in;}

199

ol

200

{margin-bottom:0in;}

201

ul

202

{margin-bottom:0in;}

203

-->

204

</style>

205

<!--[if gte mso 10]>

206

<style>

207

/* Style Definitions */

208

table.MsoNormalTable

209

{mso-style-name:"Table Normal";

210

mso-tstyle-rowband-size:0;

211

mso-tstyle-colband-size:0;

212

mso-style-noshow:yes;

213

mso-style-parent:"";

214

mso-padding-alt:0in 5.4pt 0in 5.4pt;

215

mso-para-margin:0in;

216

mso-para-margin-bottom:.0001pt;

217

mso-pagination:widow-orphan;

218

font-size:10.0pt;

219

font-family:"Times New Roman";

220

mso-ansi-language:#0400;

221

mso-fareast-language:#0400;

222

mso-bidi-language:#0400;}

223

</style>

224

<![endif]--><!--[if gte mso 9]><xml>

225

<o:shapedefaults v:ext="edit" spidmax="2050"/>

226

</xml><![endif]--><!--[if gte mso 9]><xml>

227

<o:shapelayout v:ext="edit">

228

<o:idmap v:ext="edit" data="1"/>

229

</o:shapelayout></xml><![endif]-->

230

</head>

231

232

233

234

235

236

<span style='mso-ansi-language:

237

EN-US'><o:p> </o:p>

238

239

<span style='mso-ansi-language:

240

EN-US'><o:p> </o:p>

241

242

<span style='mso-ansi-language:

243

EN-US'><o:p> </o:p>

244

245

<span style='mso-ansi-language:

246

EN-US'><o:p> </o:p>

247

248

<span style='mso-ansi-language:

249

EN-US'><o:p> </o:p>

250

251

<span style='mso-ansi-language:

252

EN-US'><o:p> </o:p>

253

254

<span style='mso-ansi-language:

255

EN-US'><o:p> </o:p>

256

257

<span style='mso-ansi-language:

258

EN-US'><o:p> </o:p>

259

260

<span style='mso-ansi-language:

261

EN-US'><o:p> </o:p>

262

263

<span style='mso-ansi-language:

264

EN-US'><o:p> </o:p>

265

266

<span style='mso-ansi-language:

267

EN-US'><o:p> </o:p>

268

269

<span style='mso-ansi-language:

270

EN-US'><o:p> </o:p>

271

272

<span style='mso-ansi-language:

273

EN-US'><o:p> </o:p>

274

275

<span style='mso-ansi-language:

276

EN-US'><o:p> </o:p>

277

278

<span style='mso-ansi-language:

279

EN-US'><o:p> </o:p>

280

281

<span style='mso-ansi-language:

282

EN-US'><o:p> </o:p>

283

284

<span style='mso-ansi-language:

285

EN-US'><o:p> </o:p>

286

287

<span

288

style='mso-ansi-language:EN-US'><o:p> </o:p>

289

290

<span

291

style='font-size:24.0pt;mso-ansi-language:EN-US'>Digital Image Correlation<o:p></o:p>

292

293

<span

294

style='font-size:24.0pt;mso-ansi-language:EN-US'>and Tracking with Matlab<o:p></o:p>

295

296

<span style='mso-ansi-language:

297

EN-US'><o:p> </o:p>

298

299

<span style='mso-ansi-language:

300

EN-US'><o:p> </o:p>

301

302

<span

303

style='mso-ansi-language:EN-US'><o:p> </o:p>

304

305

<span style='mso-ansi-language:

306

EN-US'><o:p> </o:p>

307

308

<span style='mso-ansi-language:

309

EN-US'><o:p> </o:p>

310

311

<span style='mso-ansi-language:

312

EN-US'><o:p> </o:p>

313

314

<span style='mso-ansi-language:

315

EN-US'><o:p> </o:p>

316

317

<span style='mso-ansi-language:

318

EN-US'><o:p> </o:p>

319

320

<span style='mso-ansi-language:

321

EN-US'><o:p> </o:p>

322

323

<span style='mso-ansi-language:

324

EN-US'><o:p> </o:p>

325

326

<span style='mso-ansi-language:

327

EN-US'><o:p> </o:p>

328

329

<span style='mso-ansi-language:

330

EN-US'><o:p> </o:p>

331

332

<span style='mso-ansi-language:

333

EN-US'><o:p> </o:p>

334

335

<span style='mso-ansi-language:

336

EN-US'><o:p> </o:p>

337

338

<span style='mso-ansi-language:

339

EN-US'><o:p> </o:p>

340

341

<span style='mso-ansi-language:

342

EN-US'><o:p> </o:p>

343

344

<span style='mso-ansi-language:

345

EN-US'><o:p> </o:p>

346

347

<span style='mso-ansi-language:

348

EN-US'><o:p> </o:p>

349

350

<span style='mso-ansi-language:

351

EN-US'><o:p> </o:p>

352

353

<span style='mso-ansi-language:

354

EN-US'><o:p> </o:p>

355

356

<span style='mso-ansi-language:

357

EN-US'><o:p> </o:p>

358

359

<span style='mso-ansi-language:

360

EN-US'><o:p> </o:p>

361

362

<span

363

style='mso-ansi-language:EN-US'>Programmed by:<o:p></o:p>

364

365

<span

366

style='mso-ansi-language:EN-US'><o:p> </o:p>

367

368

<span

369

style='mso-ansi-language:EN-US'>Christoph Eberl, Robert Thompson, Daniel

370

Gianola<o:p></o:p>

371

372

<span

373

style='mso-ansi-language:EN-US'>@ <st1:place w:st="on"><st1:PlaceName w:st="on">Johns</st1:PlaceName>

374

<st1:PlaceName w:st="on">Hopkins</st1:PlaceName> <st1:PlaceType w:st="on">University</st1:PlaceType></st1:place>,

375

Group of Kevin J. Hemker<o:p></o:p>

376

377

<span style='mso-ansi-language:

378

EN-US'><o:p> </o:p>

379

380

<span

381

style='mso-ansi-language:EN-US'><a href="mailto:Chris.eberl@jhu.edu">chris.eberl@jhu.edu</a><o:p></o:p>

382

383

<span style='font-size:12.0pt;font-family:"Times New Roman";mso-fareast-font-family:

384

"Times New Roman";mso-ansi-language:EN-US;mso-fareast-language:DE;mso-bidi-language:

385

AR-SA'>

386

387

388

<span style='font-size:18.0pt;

389

mso-ansi-language:EN-US'>1. Introduction<o:p></o:p>

390

391

<span style='mso-ansi-language:

392

EN-US'><o:p> </o:p>

393

394

<span style='mso-ansi-language:

395

EN-US'>Measuring strain in samples which are too small, big, compliant, soft or

396

hot are typical scenarios where non-contact techniques are needed. A technique

397

which can cover all that and also can deal with complicated strain fields in

398

structures or structural materials is the Digital Image Correlation. With this

399

technique, strain can be calculated from a series of consecutive images with

400

sub pixel resolution as will be shown in the following chapters. <o:p></o:p>

401

402

<span style='mso-ansi-language:

403

EN-US'><o:p> </o:p>

404

405

<span style='mso-ansi-language:

406

EN-US'>Even though there are tons of codes from the image registration, artificial

407

intelligence or the robotics community, none of them can easily be used by the

408

strain measuring community. Commercial code is available also and has the

409

advantage of getting a guaranty that it works, is nicely designed and has well

410

thought through user interfaces and typically a higher processing speed. The

411

disadvantages are, that commercial software typically has to be paid in k$, is available

412

only as package with hardware, enjoys a notorious lack of programming interfaces

413

or tools to change the code to fit it into a test setup as well as the

414

probability of inaccessible data in case the software license is not valid

415

anymore or it does not run on the new and fancy computer anymore. <o:p></o:p>

416

417

<span style='mso-ansi-language:

418

EN-US'><o:p> </o:p>

419

420

<span style='mso-ansi-language:

421

EN-US'>Out of all these reasons this code was written together with Rob

422

Thompson and Dan Gianola during my stay in the group of Kevin J. Hemker at the <st1:PlaceName

423

w:st="on">Johns</st1:PlaceName> <st1:PlaceName w:st="on">Hopkins</st1:PlaceName>

424

<st1:PlaceType w:st="on">University</st1:PlaceType> in <st1:place w:st="on"><st1:City

425

w:st="on">Baltimore</st1:City>, <st1:State w:st="on">MD</st1:State>, <st1:country-region

426

w:st="on">USA</st1:country-region></st1:place>. This code is not meant to be a

427

direct competitor to commercial code since we have not the time to make it as

428

easy to use as possible but as a different option with the advantages to be �free�,

429

�flexible� and �scalable�. �Free� in terms of free access even though we would like

430

to ask you to cite our code in case you use it and �free� again even though you

431

need to buy matlab together with some toolboxes. Since most research

432

institutions have access to this important tool I think we still can name it

433

�free�. �Flexible� in terms of the relative easy way you can enhance this

434

matlab code as a script language where you can add either other toolboxes or

435

your own code to flex it around your application. We would appreciate it if you

436

as a user could share your own code with all of us out here so we can learn

437

from your creativity. And �scalable� since you can easily start several

438

sessions to process your images on more than one processor (core) and because

439

there is a good chance that we will be able to use Graphic Processing Units

440

(GPU = the graphics processor on a graphics card) or other add-on boards to

441

enhance processing speed in the next few years.<o:p></o:p>

442

443

<span style='mso-ansi-language:

444

EN-US'><o:p> </o:p>

445

446

<span style='mso-ansi-language:

447

EN-US'>In case you are still reading, we would like to wish you fun using this

448

code and hope we were able to provide you with a useful tool to help your with

449

your experiments. <o:p></o:p>

450

451

<span style='mso-ansi-language:

452

EN-US'><o:p> </o:p>

453

454

<span style='mso-ansi-language:

455

EN-US'>Cheers, Chris. <o:p></o:p>

456

457

<span style='mso-ansi-language:

458

EN-US'><o:p> </o:p>

459

460

<st1:country-region w:st="on"><st1:place

461

w:st="on">Germany</st1:place></st1:country-region><span

462

style='mso-ansi-language:EN-US'>, December 2006.<o:p></o:p>

463

464

<span style='mso-ansi-language:

465

EN-US'><o:p> </o:p>

466

467

<span style='font-size:12.0pt;font-family:"Times New Roman";mso-fareast-font-family:

468

"Times New Roman";mso-ansi-language:EN-US;mso-fareast-language:DE;mso-bidi-language:

469

AR-SA'>

470

471

472

<span style='font-size:18.0pt;

473

mso-ansi-language:EN-US'>2. Requierements and Installation<o:p></o:p>

474

475

<span style='mso-ansi-language:

476

EN-US'><o:p> </o:p>

477

478

<span style='mso-ansi-language:

479

EN-US'><o:p> </o:p>

480

481

<b style='mso-bidi-font-weight:

482

normal'>REQUIEREMENTS:<o:p></o:p>

483

484

<span style='mso-ansi-language:

485

EN-US'><o:p> </o:p>

486

487

<span style='mso-ansi-language:

488

EN-US'>You will need Matlab 7 (R14) or

489

higher (since the dlmwrite.m in Matlab 6.5 does not work, at least for me

490

and this is a really important function since it stores the data after each

491

calculation step). And you will need the following <b style='mso-bidi-font-weight:

492

normal'>TOOLBOXES:<o:p></o:p>

493

494

<span style='mso-ansi-language:

495

EN-US'><o:p> </o:p>

496

497

<p class=MsoNormal style='margin-left:.5in;text-align:justify;text-indent:-.25in;

498

mso-list:l1 level1 lfo3;tab-stops:list .5in'><![if !supportLists]><span

499

style='mso-ansi-language:EN-US'>-<span

500

style='font:7.0pt "Times New Roman"'>         

501

<![endif]><span

502

style='mso-ansi-language:EN-US'>Optimization<span style='mso-ansi-language:

503

EN-US'> (all fitting processes depends on this toolbox)<o:p></o:p>

504

505

<p class=MsoNormal style='margin-left:.5in;text-align:justify;text-indent:-.25in;

506

mso-list:l1 level1 lfo3;tab-stops:list .5in'><![if !supportLists]><span

507

style='mso-ansi-language:EN-US'>-<span

508

style='font:7.0pt "Times New Roman"'>         

509

<![endif]><span

510

style='mso-ansi-language:EN-US'>Image processing<span

511

style='mso-ansi-language:EN-US'> (obviously)<o:p></o:p>

512

513

<span style='mso-ansi-language:

514

EN-US'><o:p> </o:p>

515

516

<span style='mso-ansi-language:

517

EN-US'><o:p> </o:p>

518

519

<b style='mso-bidi-font-weight:

520

normal'>INSTALLATION STEP 1:<o:p></o:p>

521

522

<span style='mso-ansi-language:

523

EN-US'><o:p> </o:p>

524

525

<span style='mso-ansi-language:

526

EN-US'>Copy the following essential

527

files into the work folder in your matlab folder (e.g. in windows:

528

c:\matlab65\work):<o:p></o:p>

529

530

<span style='mso-ansi-language:

531

EN-US'><o:p> </o:p>

532

533

<p class=MsoNormal style='margin-left:.5in;text-align:justify;text-indent:-.25in;

534

mso-list:l2 level1 lfo2;tab-stops:list .5in'><![if !supportLists]><span

535

style='mso-ansi-language:EN-US'>-<span

536

style='font:7.0pt "Times New Roman"'>         

537

<![endif]>filelist_generator.m<span

538

style='mso-tab-count:1'>�� (generates file name lists and an optional time_image

539

<o:p></o:p>

540

541

<p class=MsoNormal style='margin-left:1.75in;text-align:justify;text-indent:

542

15.6pt'>list needed for merging stress

543

and strain)<o:p></o:p>

544

545

<p class=MsoNormal style='margin-left:.5in;text-align:justify;text-indent:-.25in;

546

mso-list:l2 level1 lfo2;tab-stops:list .5in'><![if !supportLists]><span

547

style='mso-ansi-language:EN-US'>-<span

548

style='font:7.0pt "Times New Roman"'>         

549

<![endif]>grid_generator.m<span

550

style='mso-tab-count:1'>�� (generates grid rasters needed for the

551

correlation code)<o:p></o:p>

552

553

<p class=MsoNormal style='margin-left:.5in;text-align:justify;text-indent:-.25in;

554

mso-list:l2 level1 lfo2;tab-stops:list .5in'><![if !supportLists]><span

555

style='mso-ansi-language:EN-US'>-<span

556

style='font:7.0pt "Times New Roman"'>         

557

<![endif]>automate_image<span

558

style='mso-tab-count:1'>�� (this function does all the hard

559

correlation work)<o:p></o:p>

560

561

<p class=MsoNormal style='margin-left:.5in;text-align:justify;text-indent:-.25in;

562

mso-list:l2 level1 lfo2;tab-stops:list .5in'><![if !supportLists]><span

563

style='mso-ansi-language:EN-US'>-<span

564

style='font:7.0pt "Times New Roman"'>         

565

<![endif]>peak_labelling.m<span

566

style='mso-tab-count:1'>�� (this function is searching and tracking

567

peaks)<o:p></o:p>

568

569

<p class=MsoNormal style='margin-left:.5in;text-align:justify;text-indent:-.25in;

570

mso-list:l2 level1 lfo2;tab-stops:list .5in'><![if !supportLists]><span

571

style='mso-ansi-language:EN-US'>-<span

572

style='font:7.0pt "Times New Roman"'>         

573

<![endif]>pickpeak.m<span

574

style='mso-tab-count:2'>�� (this function is tracking

575

manually chosen peaks)<o:p></o:p>

576

577

<p class=MsoNormal style='margin-left:.5in;text-align:justify;text-indent:-.25in;

578

mso-list:l2 level1 lfo2;tab-stops:list .5in'><![if !supportLists]><span

579

style='mso-ansi-language:EN-US'>-<span

580

style='font:7.0pt "Times New Roman"'>         

581

<![endif]>strain_lineprofile.m<span

582

style='mso-tab-count:1'>�� (tracking two markers in a lineprofile)<o:p></o:p>

583

584

<p class=MsoNormal style='margin-left:.5in;text-align:justify;text-indent:-.25in;

585

mso-list:l2 level1 lfo2;tab-stops:list .5in'><![if !supportLists]><span

586

style='mso-ansi-language:EN-US'>-<span

587

style='font:7.0pt "Times New Roman"'>         

588

<![endif]>line_visuals.m<span

589

style='mso-tab-count:2'>�� (needed for the strain_lineprofile.m)<o:p></o:p>

590

591

<p class=MsoNormal style='margin-left:.5in;text-align:justify;text-indent:-.25in;

592

mso-list:l2 level1 lfo2;tab-stops:list .5in'><![if !supportLists]><span

593

style='mso-ansi-language:EN-US'>-<span

594

style='font:7.0pt "Times New Roman"'>         

595

<![endif]>sortvalidpoints.m<span

596

style='mso-tab-count:1'>�� (this function finds the tracked peaks

597

and has to be �� called

598

after peak_labelling or pickpeak)<o:p></o:p>

599

600

<p class=MsoNormal style='margin-left:.5in;text-align:justify;text-indent:-.25in;

601

mso-list:l2 level1 lfo2;tab-stops:list .5in'><![if !supportLists]><span

602

style='mso-ansi-language:EN-US'>-<span

603

style='font:7.0pt "Times New Roman"'>         

604

<![endif]>gauss_onepk.m<span

605

style='mso-tab-count:1'>�� (the gauss equation called by the

606

peaktracking functions)<o:p></o:p>

607

608

<p class=MsoNormal style='margin-left:.5in;text-align:justify;text-indent:-.25in;

609

mso-list:l2 level1 lfo2;tab-stops:list .5in'><![if !supportLists]><span

610

style='mso-ansi-language:EN-US'>-<span

611

style='font:7.0pt "Times New Roman"'>         

612

<![endif]>gauss_twopk.m<span

613

style='mso-tab-count:1'>�� (same as gauss_onepk.m but with two

614

peaks�)<o:p></o:p>

615

616

<p class=MsoNormal style='margin-left:.5in;text-align:justify;text-indent:-.25in;

617

mso-list:l2 level1 lfo2;tab-stops:list .5in'><![if !supportLists]><span

618

style='mso-ansi-language:EN-US'>-<span

619

style='font:7.0pt "Times New Roman"'>         

620

<![endif]>displacement.m<span

621

style='mso-tab-count:1'>�� (this function will help you analyzing

622

your data)<o:p></o:p>

623

624

<p class=MsoNormal style='margin-left:.5in;text-align:justify;text-indent:-.25in;

625

mso-list:l2 level1 lfo2;tab-stops:list .5in'><![if !supportLists]><span

626

style='mso-ansi-language:EN-US'>-<span

627

style='font:7.0pt "Times New Roman"'>         

628

<![endif]>linearfit.m<span

629

style='mso-tab-count:2'>�� (contains the linear

630

equation)<o:p></o:p>

631

632

<p class=MsoNormal style='margin-left:.5in;text-align:justify;text-indent:-.25in;

633

mso-list:l2 level1 lfo2;tab-stops:list .5in'><![if !supportLists]><span

634

style='mso-ansi-language:EN-US'>-<span

635

style='font:7.0pt "Times New Roman"'>         

636

<![endif]>ppselection_func.m<span

637

style='mso-tab-count:1'>�� (this function is needed by displacement.m)<o:p></o:p>

638

639

<p class=MsoNormal style='margin-left:.5in;text-align:justify;text-indent:-.25in;

640

mso-list:l2 level1 lfo2;tab-stops:list .5in'><![if !supportLists]><span

641

style='mso-ansi-language:EN-US'>-<span

642

style='font:7.0pt "Times New Roman"'>         

643

<![endif]>resume_automate_image.m

644

(resume stopped correlation jobs, see Chapt. 6)<o:p></o:p>

645

646

<p class=MsoNormal style='margin-left:.5in;text-align:justify;text-indent:-.25in;

647

mso-list:l2 level1 lfo2;tab-stops:list .5in'><![if !supportLists]><span

648

style='mso-ansi-language:EN-US'>-<span

649

style='font:7.0pt "Times New Roman"'>         

650

<![endif]>jobskript.m<span

651

style='mso-tab-count:2'>�� (generates a pile of jobs and

652

executes them Chapt. 6)<o:p></o:p>

653

654

<span style='mso-ansi-language:

655

EN-US'><o:p> </o:p>

656

657

<span style='mso-ansi-language:

658

EN-US'><o:p> </o:p>

659

660

<b style='mso-bidi-font-weight:

661

normal'>INSTALLATION STEP 2:<o:p></o:p>

662

663

<span style='mso-ansi-language:

664

EN-US'><o:p> </o:p>

665

666

<span style='mso-ansi-language:

667

EN-US'>In cpcorr.m (type �open cpcorr� at the matlab prompt) you have to change

668

<o:p></o:p>

669

670

<span style='mso-ansi-language:

671

EN-US'><o:p> </o:p>

672

673

<p class=MsoNormal style='margin-left:.25in;text-align:justify;text-indent:

674

-.25in;mso-list:l2 level1 lfo2;tab-stops:list .25in'><![if !supportLists]><span

675

style='mso-ansi-language:EN-US'>-<span

676

style='font:7.0pt "Times New Roman"'>         

677

<![endif]><span

678

style='mso-ansi-language:EN-US'>in line 77: <o:p></o:p>

679

680

<b style='mso-bidi-font-weight:

681

normal'><o:p> </o:p>

682

683

<span

684

style='mso-ansi-language:EN-US'>CORRSIZE = 5; <o:p></o:p>

685

686

<span

687

style='mso-ansi-language:EN-US'>�� to: <o:p></o:p>

688

689

<span

690

style='mso-ansi-language:EN-US'>CORRSIZE = 15;<o:p></o:p>

691

692

<span

693

style='mso-ansi-language:EN-US'><o:p> </o:p>

694

695

<i

696

style='mso-bidi-font-style:normal'>(This

697

changes the size of the selected parts of the image which will be correlated

698

from 10x10 pixels to 30x30. Change this to smaller values if you experience

699

slow computational speed or if you use low resolution images. Remember that

700

markers need more than double the space from its centre to the edge of the

701

image, otherwise they cannot be tracked.)<o:p></o:p>

702

703

<span

704

style='mso-ansi-language:EN-US'><o:p> </o:p>

705

706

<p class=MsoNormal style='margin-left:.25in;text-align:justify;text-indent:

707

-.25in;mso-list:l2 level1 lfo2;tab-stops:list .25in'><![if !supportLists]><span

708

style='mso-ansi-language:EN-US'>-<span

709

style='font:7.0pt "Times New Roman"'>         

710

<![endif]><span

711

style='mso-ansi-language:EN-US'>in line 134 and 135:<o:p></o:p>

712

713

<span

714

style='mso-ansi-language:EN-US'><o:p> </o:p>

715

716

<span

717

style='mso-ansi-language:EN-US'>input_fractional_offset = xyinput(icp,:) -

718

round(xyinput(icp,:));<o:p></o:p>

719

720

<span

721

style='mso-ansi-language:EN-US'>base_fractional_offset = xybase_in(icp,:) - round(xybase_in(icp,:));<span

722

style='mso-spacerun:yes'>�� <o:p></o:p>

723

724

<span

725

style='mso-ansi-language:EN-US'>�� to:<o:p></o:p>

726

727

<span

728

style='mso-ansi-language:EN-US'>input_fractional_offset = xyinput(icp,:) -

729

round(xyinput(icp,:)*1000)/1000;<o:p></o:p>

730

731

<span

732

style='mso-ansi-language:EN-US'>base_fractional_offset = xybase_in(icp,:) -

733

round(xybase_in(icp,:)*1000)/1000;�� <o:p></o:p>

734

735

<span style='mso-ansi-language:

736

EN-US'><o:p> </o:p>

737

738

<i

739

style='mso-bidi-font-style:normal'>(This

740

is changing the resolution of the marker positions to 1/1000th

741

pixel. If you need higher resolution just increase these values)<o:p></o:p>

742

743

<span style='mso-ansi-language:

744

EN-US'><o:p> </o:p>

745

746

<span style='mso-ansi-language:

747

EN-US'><o:p> </o:p>

748

749

<span style='font-size:12.0pt;font-family:"Times New Roman";mso-fareast-font-family:

750

"Times New Roman";mso-ansi-language:EN-US;mso-fareast-language:DE;mso-bidi-language:

751

AR-SA'>

752

753

754

<span style='font-size:18.0pt;

755

mso-ansi-language:EN-US'>3. Good things to know about matlab:<o:p></o:p>

756

757

<span style='mso-ansi-language:

758

EN-US'><o:p> </o:p>

759

760

<span style='mso-ansi-language:

761

EN-US'>The matlab help is extremely helpful and should be the very first location

762

you look in case of errors. If you never worked with matlab but would like to

763

check it out, the �Getting Started� is a good point to start with. I started

764

there in summer 2005. <o:p></o:p>

765

766

<span style='mso-ansi-language:

767

EN-US'><o:p> </o:p>

768

769

<b style='mso-bidi-font-weight:

770

normal'>TAB:<o:p></o:p>

771

772

<span style='mso-ansi-language:

773

EN-US'>Pressing the TAB key on your keyboard after you started typing in a

774

command at the command line of matlab will show you all functions with the same

775

first letters.<o:p></o:p>

776

777

<span style='mso-ansi-language:

778

EN-US'><o:p> </o:p>

779

780

<b style='mso-bidi-font-weight:

781

normal'>Arrow up function:<o:p></o:p>

782

783

<span style='mso-ansi-language:

784

EN-US'>Pressing the Arrow Up key on your keyboard after you started typing a

785

command will show the last command

786

you started with the same first letters.<o:p></o:p>

787

788

<span style='mso-ansi-language:

789

EN-US'><o:p> </o:p>

790

791

<b style='mso-bidi-font-weight:

792

normal'>Current Folder:<o:p></o:p>

793

794

<span style='mso-ansi-language:

795

EN-US'>The �Current Folder� of matlab is the folder on your harddisk which is

796

currently selected to process data in matlab (close to the upper edge of the

797

matlab window). Functions like �automate_image.m� require certain files to be

798

present in the �Current Folder� otherwise they will produce an error (see

799

description later on). Pressing the little button on the right hand side with

800

the three little dots on it will let you select another folder. Another

801

possibility is to use the command window (see matlab help) or you can select

802

the current directory if it is activated under �View� in that extra window.<o:p></o:p>

803

804

<span style='mso-ansi-language:

805

EN-US'><o:p> </o:p>

806

807

<b style='mso-bidi-font-weight:

808

normal'>Set semicolon:<o:p></o:p>

809

810

<span style='mso-ansi-language:

811

EN-US'>Set the semicolon after calling a function (e.g. �automate_image;�),

812

otherwise all data which you get back from a called function will be plotted in

813

the command window.<o:p></o:p>

814

815

<span style='mso-ansi-language:

816

EN-US'><o:p> </o:p>

817

818

<b style='mso-bidi-font-weight:

819

normal'>Workspace:<o:p></o:p>

820

821

<span style='mso-ansi-language:

822

EN-US'>The Workspace is the place where you can load all your data into.

823

Functions called by you will write their values into the workspace and scripts

824

will use the workspace all the time and leave a mess of variables in there. If

825

you do not know what is going on check out the �Getting Started� paragraph in

826

the matlab help. <o:p></o:p>

827

828

<span style='mso-ansi-language:

829

EN-US'><o:p> </o:p>

830

831

<b style='mso-bidi-font-weight:

832

normal'>How to load data?<o:p></o:p>

833

834

<span style='mso-ansi-language:

835

EN-US'>If you have loaded data into the workspace (either by choosing �File� <span

836

style='font-family:Wingdings;mso-ascii-font-family:"Times New Roman";

837

mso-hansi-font-family:"Times New Roman";mso-ansi-language:EN-US;mso-char-type:

838

symbol;mso-symbol-font-family:Wingdings'><span style='mso-char-type:symbol;

839

mso-symbol-font-family:Wingdings'>�<span style='mso-ansi-language:

840

EN-US'> �Open�� and selected the data file you wanted to load or using the

841

command window e.g. by typing: �load('filenamelist')� and filenamelist is

842

present in the Current Directory) the data will appear in the workspace window.<o:p></o:p>

843

844

<span style='mso-ansi-language:

845

EN-US'><o:p> </o:p>

846

847

<b style='mso-bidi-font-weight:

848

normal'>How to saved data from Workspace

849

to the hard disk?<o:p></o:p>

850

851

<span style='mso-ansi-language:

852

EN-US'>If you want to save data from the workspace to the hard disk, right

853

click on it and select �Save Selection As��. It will save the data with the

854

matlab file fomat. This data is only accessible by matlab. If you want to

855

process the data also with other programs you should consider to save the data

856

as ASCII file. Therefore type in the matlab console window �save('stress.txt','alltemp','-ASCII');�

857

to save the variable alltemp as text file�

858

with the name �stress.txt� to the �Current directory�.<span

859

style='mso-spacerun:yes'>� If you want to open it with matlab or excel

860

you have to import the file. The delimiter is per default TAB but can also be

861

chosen to be comma or space (see matlab help).<o:p></o:p>

862

863

<span style='mso-ansi-language:

864

EN-US'><o:p> </o:p>

865

866

<b style='mso-bidi-font-weight:

867

normal'>How to give data in the workspace

868

to the functions?<o:p></o:p>

869

870

<span style='mso-ansi-language:

871

EN-US'>If you type �displacement;� into the command window of matlab, the

872

function will start and ask you for the needed files. Instead you can load the

873

data (e.g. validx.mat and validy.mat) into the workspace and then give it to

874

the displacement function by typing: �displacement(validx,validy);�. If have

875

not loaded validx.mat there will be an error message. <o:p></o:p>

876

877

<span style='mso-ansi-language:

878

EN-US'><o:p> </o:p>

879

880

<b style='mso-bidi-font-weight:

881

normal'>How do I get the data from the

882

function into the workspace?<o:p></o:p>

883

884

<span style='mso-ansi-language:

885

EN-US'>If you type �[validx,validy]=displacement;� the variables validx and validy

886

will be created after running �displacement.m�. If you have manipulated validx

887

and validy (e.g. if you cleaned up the data set from miss-tracked markers) you should

888

save them.<o:p></o:p>

889

890

<span style='mso-ansi-language:

891

EN-US'><o:p> </o:p>

892

893

<b style='mso-bidi-font-weight:

894

normal'>Why are all images<span

895

style='mso-spacerun:yes'>� opened in matlab mirrored to the horizontal center

896

axis of the image?<o:p></o:p>

897

898

<span style='mso-ansi-language:

899

EN-US'>Matlab reads in images like a diagram. Therefore pixel (1,1) is in the

900

lower left while image processing software starts at the upper left corner. <o:p></o:p>

901

902

<span style='mso-ansi-language:

903

EN-US'><o:p> </o:p>

904

905

<b style='mso-bidi-font-weight:

906

normal'>How o stop a function in matlab?<span

907

style='mso-ansi-language:EN-US'><o:p></o:p>

908

909

<span style='mso-ansi-language:

910

EN-US'>To stop a function press the control key (�Crtl�) together with �c�.<o:p></o:p>

911

912

<span style='font-size:12.0pt;font-family:"Times New Roman";mso-fareast-font-family:

913

"Times New Roman";mso-ansi-language:EN-US;mso-fareast-language:DE;mso-bidi-language:

914

AR-SA'>

915

916

917

<span style='font-size:18.0pt;

918

mso-ansi-language:EN-US'>4. Digital Image Correlation Quick Guide<o:p></o:p>

919

920

<span style='mso-ansi-language:

921

EN-US'><o:p> </o:p>

922

923

<span style='mso-ansi-language:

924

EN-US'>This guide should help you to perform a simple and fast analysis of your

925

images. Before we start you should check your image format and the naming of

926

your files. The preferred image format is *.tif and can be compressed with the

927

packbits compression. JPEG or other image formats as well as MPEG video

928

compression formats will not provide you with sub pixel resolution since the

929

images are processed to save as much space as possible. <o:p></o:p>

930

931

<span style='mso-ansi-language:

932

EN-US'>The script we use to create a list of images to process

933

(filelist_generator.m) is kind of limited to a certain format but it is

934

possible to generate your own list of images. If you want to change the format

935

or the names of your images you can use free programs like Irfanview (<a

936

href="http://www.irfanview.com/">www.irfanview.com</a>) to batch process a huge

937

number of images.<o:p></o:p>

938

939

<span style='mso-ansi-language:

940

EN-US'><o:p> </o:p>

941

942

<i style='mso-bidi-font-style:

943

normal'>4 Steps to

944

Success:<o:p></o:p>

945

946

<span style='mso-ansi-language:

947

EN-US'><o:p> </o:p>

948

949

<p class=MsoNormal style='margin-left:.5in;text-align:justify;text-indent:-.25in;

950

mso-list:l0 level1 lfo1;tab-stops:list .5in'><![if !supportLists]><span

951

style='font-size:16.0pt;mso-ansi-language:EN-US'>1.<span

952

style='font:7.0pt "Times New Roman"'>    <![endif]><span

953

style='font-size:16.0pt;mso-ansi-language:EN-US'>Step: Filename list generation

954

with filelist_generator.m<o:p></o:p>

955

956

<span

957

style='mso-ansi-language:EN-US'><o:p> </o:p>

958

959

<span

960

style='mso-ansi-language:EN-US'>Just type �filelist_generator;� and press

961

�ENTER� at the command line of matlab. <o:p></o:p>

962

963

<span

964

style='mso-ansi-language:EN-US'>There are two ways to proceed now: Either you

965

choose to �manually� type in the image numbers you want to be processed or you

966

can �automatically� generate a list of images by just pointing out the first

967

image and the function will find all images with increasing number before the

968

point in the name (e.g. �PIC00001.Tif�). If you choose �manually�, the

969

following window should appear:<o:p></o:p>

970

971

<span

972

style='mso-ansi-language:EN-US'><o:p> </o:p>

973

974

<span

975

style='mso-ansi-language:EN-US'><!--[if gte vml 1]><v:shapetype id="_x0000_t75"

976

coordsize="21600,21600" o:spt="75" o:preferrelative="t" path="m@4@5l@4@11@9@11@9@5xe"

977

filled="f" stroked="f">

978

<v:stroke joinstyle="miter"/>

979

<v:formulas>

980

<v:f eqn="if lineDrawn pixelLineWidth 0"/>

981

<v:f eqn="sum @0 1 0"/>

982

<v:f eqn="sum 0 0 @1"/>

983

<v:f eqn="prod @2 1 2"/>

984

<v:f eqn="prod @3 21600 pixelWidth"/>

985

<v:f eqn="prod @3 21600 pixelHeight"/>

986

<v:f eqn="sum @0 0 1"/>

987

<v:f eqn="prod @6 1 2"/>

988

<v:f eqn="prod @7 21600 pixelWidth"/>

989

<v:f eqn="sum @8 21600 0"/>

990

<v:f eqn="prod @7 21600 pixelHeight"/>

991

<v:f eqn="sum @10 21600 0"/>

992

</v:formulas>

993

<v:path o:extrusionok="f" gradientshapeok="t" o:connecttype="rect"/>

994

<o:lock v:ext="edit" aspectratio="t"/>

995

</v:shapetype><v:shape id="_x0000_i1025" type="#_x0000_t75" style='width:243pt;

996

height:102.75pt'>

997

<v:imagedata src="Correlation_Guide_2_files/image001.jpg" o:title="filelist_generator"

998

cropbottom="43901f" cropright="27208f"/>

999

</v:shape><![endif]--><![if !vml]><img border=0 width=324 height=137

1000

src="Correlation_Guide_2_files/image002.jpg" v:shapes="_x0000_i1025"><![endif]><o:p></o:p>

1001

1002

<span

1003

style='mso-ansi-language:EN-US'><o:p> </o:p>

1004

1005

<span style='mso-ansi-language:

1006

EN-US'>Fig. 1: Input of first and last image to create an image list with

1007

filelist_generator.m<o:p></o:p>

1008

1009

<span

1010

style='mso-ansi-language:EN-US'><o:p> </o:p>

1011

1012

<span

1013

style='mso-ansi-language:EN-US'>The numbers will be the number at the end of

1014

each filename. After depositing these numbers in the dialog the next window

1015

will ask for the first 4 letters of the filenames.<o:p></o:p>

1016

1017

<span

1018

style='mso-ansi-language:EN-US'><o:p> </o:p>

1019

1020

<span

1021

style='mso-ansi-language:EN-US'><!--[if gte vml 1]><v:shape id="_x0000_i1026"

1022

type="#_x0000_t75" style='width:243pt;height:1in'>

1023

<v:imagedata src="Correlation_Guide_2_files/image003.jpg" o:title="filelist_generator_2"

1024

cropbottom="50376f" cropright="27208f"/>

1025

</v:shape><![endif]--><![if !vml]><img border=0 width=324 height=96

1026

src="Correlation_Guide_2_files/image004.jpg" v:shapes="_x0000_i1026"><![endif]><o:p></o:p>

1027

1028

<span

1029

style='mso-ansi-language:EN-US'><o:p> </o:p>

1030

1031

<span style='mso-ansi-language:

1032

EN-US'>Fig. 2: Input for the first 4 letters in filelist_generator.m<o:p></o:p>

1033

1034

<span

1035

style='mso-ansi-language:EN-US'><o:p> </o:p>

1036

1037

<span

1038

style='mso-ansi-language:EN-US'>The next step is to save the file name list

1039

into the folder with the images to process. You should choose to use

1040

�filenamelist.mat� since this is what the following scripts will search for. <o:p></o:p>

1041

1042

<span

1043

style='mso-ansi-language:EN-US'><o:p> </o:p>

1044

1045

<span

1046

style='mso-ansi-language:EN-US'><!--[if gte vml 1]><v:shape id="_x0000_i1027"

1047

type="#_x0000_t75" style='width:252pt;height:162pt'>

1048

<v:imagedata src="Correlation_Guide_2_files/image005.jpg" o:title="filelist_generator_3"

1049

cropbottom="31426f" cropright="25789f"/>

1050

</v:shape><![endif]--><![if !vml]><img border=0 width=336 height=216

1051

src="Correlation_Guide_2_files/image006.jpg" v:shapes="_x0000_i1027"><![endif]><o:p></o:p>

1052

1053

<span

1054

style='mso-ansi-language:EN-US'><o:p> </o:p>

1055

1056

<span style='mso-ansi-language:

1057

EN-US'>Fig. 3: Dialog to save the file name list into the folder with the

1058

images to analyze.<o:p></o:p>

1059

1060

<span

1061

style='mso-ansi-language:EN-US'><o:p> </o:p>

1062

1063

<span

1064

style='mso-ansi-language:EN-US'><o:p> </o:p>

1065

1066

<span

1067

style='mso-ansi-language:EN-US'><o:p> </o:p>

1068

1069

<p class=MsoNormal style='margin-left:.5in;text-align:justify;text-indent:-.25in;

1070

mso-list:l0 level1 lfo1;tab-stops:list .5in'><![if !supportLists]><span

1071

style='font-size:16.0pt;mso-ansi-language:EN-US'>2.<span

1072

style='font:7.0pt "Times New Roman"'>    <![endif]><span

1073

style='font-size:16.0pt;mso-ansi-language:EN-US'>Grid generation with grid_generator.m

1074

for correlation:<o:p></o:p>

1075

1076

<span

1077

style='mso-ansi-language:EN-US'><o:p> </o:p>

1078

1079

<span

1080

style='mso-ansi-language:EN-US'>It has to be noted that the user can always

1081

generate his own marker positions. Therefore the marker position in pixel has

1082

to be saved as a text based format where the x-position is saved as

1083

�grid_x.dat� and the y-position saved as �grid_y.dat�. <o:p></o:p>

1084

1085

<span

1086

style='mso-ansi-language:EN-US'>To start just type �grid_generator;� and press

1087

�ENTER� at the command line of matlab. The following window should appear:<o:p></o:p>

1088

1089

<span

1090

style='mso-ansi-language:EN-US'><o:p> </o:p>

1091

1092

<span

1093

style='mso-ansi-language:EN-US'><!--[if gte vml 1]><v:shape id="_x0000_i1028"

1094

type="#_x0000_t75" style='width:252pt;height:164.25pt'>

1095

<v:imagedata src="Correlation_Guide_2_files/image007.jpg" o:title="grid_generator_1"

1096

cropbottom="30952f" cropright="25789f"/>

1097

</v:shape><![endif]--><![if !vml]><img border=0 width=336 height=219

1098

src="Correlation_Guide_2_files/image008.jpg" v:shapes="_x0000_i1028"><![endif]><o:p></o:p>

1099

1100

<span

1101

style='mso-ansi-language:EN-US'><o:p> </o:p>

1102

1103

<span style='mso-ansi-language:

1104

EN-US'>Fig. 4: Dialog to open the first (base) image to generate a grid<o:p></o:p>

1105

1106

<span

1107

style='mso-ansi-language:EN-US'><o:p> </o:p>

1108

1109

<span

1110

style='mso-ansi-language:EN-US'>In this dialog the first (base) image can be

1111

selected in which the grid can be created. After selecting this base image, the

1112

image will be opened and a new dialog pops up to ask you what kind of grid

1113

should be used.<o:p></o:p>

1114

1115

<span

1116

style='mso-ansi-language:EN-US'><o:p> </o:p>

1117

1118

<span

1119

style='mso-ansi-language:EN-US'><!--[if gte vml 1]><v:shape id="_x0000_i1030"

1120

type="#_x0000_t75" style='width:279pt;height:135pt'>

1121

<v:imagedata src="Correlation_Guide_2_files/image009.jpg" o:title="grid_generator_3"/>

1122

</v:shape><![endif]--><![if !vml]><img border=0 width=372 height=180

1123

src="Correlation_Guide_2_files/image010.jpg" v:shapes="_x0000_i1030"><![endif]><!--[if gte vml 1]><v:shape

1124

id="_x0000_i1031" type="#_x0000_t75" style='width:99pt;height:132.75pt'>

1125

<v:imagedata src="Correlation_Guide_2_files/image011.jpg" o:title="grid_generator_2"

1126

cropbottom="31426f" cropright="46490f"/>

1127

</v:shape><![endif]--><![if !vml]><img border=0 width=132 height=177

1128

src="Correlation_Guide_2_files/image012.jpg" v:shapes="_x0000_i1031"><![endif]><o:p></o:p>

1129

1130

<span

1131

style='mso-ansi-language:EN-US'><o:p> </o:p>

1132

1133

<span style='mso-ansi-language:

1134

EN-US'>Fig. 5: The opened base image and the menu to select the grid type.<o:p></o:p>

1135

1136

<span

1137

style='mso-ansi-language:EN-US'><o:p> </o:p>

1138

1139

<span

1140

style='mso-ansi-language:EN-US'>The different types are a rectangular or

1141

circular grid, two markers, a line or two rectangles of markers. If you choose a

1142

rectangular grid type, the pointer will change from an arrow to a horizontal

1143

and a vertical line which will help you finding the right position. The idea is

1144

to click on the two diagonal positions which will define the outer dimensions

1145

of a box containing the grid.<o:p></o:p>

1146

1147

<span

1148

style='mso-ansi-language:EN-US'><o:p> </o:p>

1149

1150

<span

1151

style='mso-ansi-language:EN-US'><!--[if gte vml 1]><v:shape id="_x0000_i1029"

1152

type="#_x0000_t75" style='width:414.75pt;height:200.25pt'>

1153

<v:imagedata src="Correlation_Guide_2_files/image013.jpg" o:title="grid_generator_4"/>

1154

</v:shape><![endif]--><![if !vml]><img border=0 width=553 height=267

1155

src="Correlation_Guide_2_files/image014.jpg" v:shapes="_x0000_i1029"><![endif]><o:p></o:p>

1156

1157

<span

1158

style='mso-ansi-language:EN-US'><o:p> </o:p>

1159

1160

<span style='mso-ansi-language:

1161

EN-US'>Fig. 6: The horizontal and vertical lines allow an accurate positioning

1162

of the grid.<o:p></o:p>

1163

1164

<span

1165

style='mso-ansi-language:EN-US'><o:p> </o:p>

1166

1167

<span

1168

style='mso-ansi-language:EN-US'>The selected box will be shown in the image and

1169

a dialog will pop up to ask your for the input of a raster point distance in x

1170

and y direction. <o:p></o:p>

1171

1172

<span

1173

style='mso-ansi-language:EN-US'><o:p> </o:p>

1174

1175

<span

1176

style='mso-ansi-language:EN-US'><!--[if gte vml 1]><v:shape id="_x0000_i1032"

1177

type="#_x0000_t75" style='width:243pt;height:99pt'>

1178

<v:imagedata src="Correlation_Guide_2_files/image015.jpg" o:title="grid_generator_5"

1179

cropbottom="44691f" cropright="27208f"/>

1180

</v:shape><![endif]--><![if !vml]><img border=0 width=324 height=132

1181

src="Correlation_Guide_2_files/image016.jpg" v:shapes="_x0000_i1032"><![endif]><o:p></o:p>

1182

1183

<span

1184

style='mso-ansi-language:EN-US'><o:p> </o:p>

1185

1186

<span style='mso-ansi-language:

1187

EN-US'>Fig. 7: Horizontal (x-direction) and vertical (y-direction) grid

1188

resolution with a default resolution of 50 pixel distance between raster points.<o:p></o:p>

1189

1190

<span

1191

style='mso-ansi-language:EN-US'><o:p> </o:p>

1192

1193

<span

1194

style='mso-ansi-language:EN-US'>The code will now generate the chosen grid and will

1195

plot it on top of the sample image. The last dialogue will ask you if you want

1196

to use the generated grid and save the grid_x.dat and grid_y.dat to be

1197

processed, if you want to try again or if you want to choose another grid type.<o:p></o:p>

1198

1199

<span

1200

style='mso-ansi-language:EN-US'><o:p> </o:p>

1201

1202

<span

1203

style='mso-ansi-language:EN-US'><!--[if gte vml 1]><v:shape id="_x0000_i1033"

1204

type="#_x0000_t75" style='width:117pt;height:97.5pt'>

1205

<v:imagedata src="Correlation_Guide_2_files/image017.jpg" o:title="grid_generator_7"

1206

cropbottom="45007f" cropright="47082f"/>

1207

</v:shape><![endif]--><![if !vml]><img border=0 width=156 height=130

1208

src="Correlation_Guide_2_files/image018.jpg" v:shapes="_x0000_i1033"><![endif]><o:p></o:p>

1209

1210

<span

1211

style='mso-ansi-language:EN-US'><o:p> </o:p>

1212

1213

<span style='mso-ansi-language:

1214

EN-US'>Fig. 8: The last menu will allow you to accept the grid (which will be

1215

saved), try again or choose another grid type.<o:p></o:p>

1216

1217

<span

1218

style='mso-ansi-language:EN-US'><o:p> </o:p>

1219

1220

<span

1221

style='mso-ansi-language:EN-US'><o:p> </o:p>

1222

1223

<p class=MsoNormal style='margin-left:.5in;text-align:justify;text-indent:-.25in;

1224

mso-list:l0 level1 lfo1;tab-stops:list .5in'><![if !supportLists]><span

1225

style='font-size:16.0pt;mso-ansi-language:EN-US'>3.<span

1226

style='font:7.0pt "Times New Roman"'>    <![endif]><span

1227

style='font-size:16.0pt;mso-ansi-language:EN-US'>Run correlation with automate_image.m:<o:p></o:p>

1228

1229

<span

1230

style='mso-ansi-language:EN-US'><o:p> </o:p>

1231

1232

<span

1233

style='mso-ansi-language:EN-US'>The automation function is the central function

1234

and processes all markers and images. Therefore the �Current directory� in

1235

matlab has to be the folder where automate_image.m finds the filenamelist.mat,

1236

grid_x.dat and grid_y.dat as well as the images specified in �filenamelist.mat�.

1237

Just type �automate_image;� and press �ENTER� at the command line of matlab. <o:p></o:p>

1238

1239

<span

1240

style='mso-ansi-language:EN-US'>At first, automate_image.m will open the first

1241

image in the filenamelist.mat and plot the grid as green crosses on top. The

1242

next step will need some time since all markers in that image have to be

1243

processed for the first image. After correlating image one and two the new

1244

raster positions will be plotted as red crosses. On top of the image and the

1245

green crosses. The next dialog will ask you if you want to continue with this

1246

correlation or cancel. If you press continue, �automate_image.m� will process

1247

all images in the �filenamelist.mat�. The time it will take to process all

1248

images will be plotted on the figure but can easily be estimated by knowing the

1249

raster point processing speed (see processing speed). <o:p></o:p>

1250

1251

<span

1252

style='mso-ansi-language:EN-US'>Depending on the number of images and markers

1253

you are tracking, this process can take between seconds and days. For 100

1254

images and 200 markers a decent computer should need 200 seconds. To get a

1255

better resolution you can always run jobs overnight (e.g. 6000 markers in 1000

1256

images) with higher resolutions. <o:p></o:p>

1257

1258

<span

1259

style='mso-ansi-language:EN-US'>Keep in mind that �CORRSIZE� which you changed

1260

in �cpcorr.m� will limit your resolution. If you chose to use the 15 pixel as

1261

suggested a marker distance of 30 pixel will lead to a full cover of the strain

1262

field. Choosing smaller marker distances will lead to an interpolation since

1263

two neighboring markers share pixels. Nevertheless a higher marker density can

1264

reduce the noise of the strain field.<o:p></o:p>

1265

1266

<span

1267

style='mso-ansi-language:EN-US'>When all images are processed, automate_image

1268

will write the files validx.mat, validy.mat, validx.txt and validy.txt. The

1269

text files are meant to store the result in a format which can be accessed by

1270

other programs also in the future. <o:p></o:p>

1271

1272

<span

1273

style='mso-ansi-language:EN-US'><o:p> </o:p>

1274

1275

<span

1276

style='mso-ansi-language:EN-US'>To stop automate_image use the key combination

1277

�Ctrl c�. <o:p></o:p>

1278

1279

<span

1280

style='mso-ansi-language:EN-US'>In case of a crashes, errors generating the

1281

right filenamelist or the interruption of the correlation process by the user,

1282

the function �recover_correlation.m� can be called which extracts the last

1283

marker positions sets up �automate_image.m� and continues at the last image.<o:p></o:p>

1284

1285

<span

1286

style='mso-ansi-language:EN-US'><o:p> </o:p>

1287

1288

<p class=MsoNormal style='margin-left:.5in;text-align:justify;text-indent:-.25in;

1289

mso-list:l0 level1 lfo1;tab-stops:list .5in'><![if !supportLists]><span

1290

style='font-size:16.0pt;mso-ansi-language:EN-US'>4.<span

1291

style='font:7.0pt "Times New Roman"'>    <![endif]><span

1292

style='font-size:16.0pt;mso-ansi-language:EN-US'>Analyze the displacement with

1293

displacement.m:<o:p></o:p>

1294

1295

<span

1296

style='mso-ansi-language:EN-US'><o:p> </o:p>

1297

1298

<span

1299

style='mso-ansi-language:EN-US'>As last part, the post processing is the most

1300

interesting and awarding step since you actually can analyze the collected

1301

displacement data. The displacement.m function is a small collection of

1302

functions which allows you to review the displacement field, calculate the

1303

strain or delete markers which were not correlated or tracked very well. <o:p></o:p>

1304

1305

<span

1306

style='mso-ansi-language:EN-US'>To start, type into the command window

1307

�displacement;� or �[validx,validy]=displacement;� in case you want to save the

1308

changed files (validx and validy) from the workspace (see chapter 3). A window

1309

will pop up asking you for the validx.dat file which contains the

1310

x-displacement off all markers in all images, followed by a dialog for the

1311

validy.dat containing the y data. After �displacement.m� has loaded both files,

1312

a new window pops up which allows you to choose between different options.<o:p></o:p>

1313

1314

<span

1315

style='mso-ansi-language:EN-US'><o:p> </o:p>

1316

1317

<span

1318

style='mso-ansi-language:EN-US'><!--[if gte vml 1]><v:shape id="_x0000_i1034"

1319

type="#_x0000_t75" style='width:179.25pt;height:241.5pt'>

1320

<v:imagedata src="Correlation_Guide_2_files/image019.jpg" o:title="displacement_1"

1321

cropbottom="14666f" cropright="37262f"/>

1322

</v:shape><![endif]--><![if !vml]><img border=0 width=239 height=322

1323

src="Correlation_Guide_2_files/image020.jpg" v:shapes="_x0000_i1034"><![endif]><o:p></o:p>

1324

1325

<span

1326

style='mso-ansi-language:EN-US'><o:p> </o:p>

1327

1328

<span style='mso-ansi-language:

1329

EN-US'>Fig. 9: The displacement.m function allows cleaning up the data set

1330

selecting parts of it plot the displacement or measure the strain in x- and y-direction.<o:p></o:p>

1331

1332

<span

1333

style='mso-ansi-language:EN-US'><o:p> </o:p>

1334

1335

<span

1336

style='mso-ansi-language:EN-US'>For a typical analysis you always have to

1337

delete some of the markers with did not all too well during the correlation or

1338

peak fitting step. This can happen e.g. due to marker movement during the test,

1339

changing light conditions or in case of the correlation technique due to the

1340

fact that the sample surface did not provide enough characteristics. <o:p></o:p>

1341

1342

<span

1343

style='mso-ansi-language:EN-US'>We start with clicking on the <b

1344

style='mso-bidi-font-weight:normal'>�3D Mesh Plot of Displacement� button which

1345

will bring up a new window and a dialog asking if you want to create a video. Clicking

1346

on �yes� will create a new folder called video and the 3D displacement plot of

1347

each image will be saved as *.jpg. The click on the no button will just start

1348

the 3D displacement plot. This part of the �displacement.m� allows you to watch

1349

displacement (z-axis) versus location (x- and y-axis) for all images. To get a

1350

better 3-dimensional understanding all markers are projected as green dots on

1351

the plane normal to the y axis. The Image number will be shown in the plot. It

1352

has to be noted that the orientation of the strain depends on the orientation

1353

of the image during the correlation process. The x-axis in the plot is the

1354

horizontal direction in the image and the y-direction the perpendicular

1355

direction. The plotted displacement on the z-axis is always the x-displacement

1356

of the data contained in validx.mat and validy.mat. To look at the y

1357

displacement the user has to wait for all images to be plotted and then after

1358

the displacement-dialog appears again, click the button �<b style='mso-bidi-font-weight:

1359

normal'>Rotate Orientation (exchange x and y)�. This will exchange validx

1360

and validy and clicking again on the �3D

1361

Mesh Plot of Displacement� button will now show the displacement in the y-direction.

1362

The user has to keep track of this change since it will affect all plotting and

1363

strain measurement steps lying ahead during the same �displacement.m� session. <o:p></o:p>

1364

1365

<span

1366

style='mso-ansi-language:EN-US'><o:p> </o:p>

1367

1368

<span

1369

style='mso-ansi-language:EN-US'><!--[if gte vml 1]><v:shape id="_x0000_i1035"

1370

type="#_x0000_t75" style='width:342pt;height:303.75pt'>

1371

<v:imagedata src="Correlation_Guide_2_files/image021.jpg" o:title="displacement_2"

1372

cropbottom="1547f" cropright="11695f"/>

1373

</v:shape><![endif]--><![if !vml]><img border=0 width=456 height=405

1374

src="Correlation_Guide_2_files/image022.jpg" v:shapes="_x0000_i1035"><![endif]><o:p></o:p>

1375

1376

<span

1377

style='mso-ansi-language:EN-US'><o:p> </o:p>

1378

1379

<span style='mso-ansi-language:

1380

EN-US'>Fig. 10: 3D Displacement versus x- and y-position. The orientation of

1381

the x-axis is the horizontal in the analyzed image and the y-axis is the

1382

vertical. The displacement is always the x-diplacement until you exchange

1383

validx and validy with the �Rotate

1384

Orientation (exchange x and y)� button.<o:p></o:p>

1385

1386

<span

1387

style='mso-ansi-language:EN-US'><o:p> </o:p>

1388

1389

<span

1390

style='mso-ansi-language:EN-US'>The next step is to get rid of badly tracked

1391

markers. This can be done in three different ways:<o:p></o:p>

1392

1393

<p class=MsoNormal style='margin-left:.5in;text-align:justify;text-indent:-.25in;

1394

mso-list:l2 level1 lfo2;tab-stops:list .5in'><![if !supportLists]><span

1395

style='mso-ansi-language:EN-US'>-<span

1396

style='font:7.0pt "Times New Roman"'>         

1397

<![endif]>delete

1398

single markers, click �Remove badly

1399

tracked marker, one by one (Position)�<o:p></o:p>

1400

1401

<p class=MsoNormal style='margin-left:.5in;text-align:justify;text-indent:-.25in;

1402

mso-list:l2 level1 lfo2;tab-stops:list .5in'><![if !supportLists]><span

1403

style='mso-ansi-language:EN-US'>-<span

1404

style='font:7.0pt "Times New Roman"'>         

1405

<![endif]>delete a

1406

bunch of markers at once, click �Delete

1407

multiple markers (Position)� <o:p></o:p>

1408

1409

<p class=MsoNormal style='margin-left:.5in;text-align:justify;text-indent:-.25in;

1410

mso-list:l2 level1 lfo2;tab-stops:list .5in'><![if !supportLists]><span

1411

style='mso-ansi-language:EN-US'>-<span

1412

style='font:7.0pt "Times New Roman"'>         

1413

<![endif]>delete a

1414

bunch of markers from the displacement versus x position plot, click �<b

1415

style='mso-bidi-font-weight:normal'>Delete markers from displacement vs.

1416

position plot�.<o:p></o:p>

1417

1418

<span

1419

style='mso-ansi-language:EN-US'>The first two will provide you with a top down

1420

view with x- and y axis being the horizontal and vertical direction in the

1421

analyzed image and the displacement is expressed as underlying colors. The third

1422

option will show the projected markers which allows you to sometimes more easily

1423

access the peaks. You have to play around with these two different views and

1424

rotate the orientation back and forth while you are checking through the images

1425

until you get a clean data set. This requires some practice, therefore take

1426

your time and analyze the data carefully before you really delete a bunch of

1427

markers.<o:p></o:p>

1428

1429

<span

1430

style='mso-ansi-language:EN-US'>�Remove

1431

badly tracked marker, one by one (Position)� will allow you to click on

1432

markers which are not at the right place. The marker with the highest

1433

displacement value will be a red dot while the marker with the lowest

1434

displacement will be a blue dot. By clicking close to one of the markers the

1435

plot will be updated to the new displacement field and the color code will be

1436

updated to the new highest and lowest displacement value.<o:p></o:p>

1437

1438

<span

1439

style='mso-ansi-language:EN-US'>�Delete

1440

multiple markers (Position)� will allow you to choose a rectangle and all

1441

markers in it will be deleted.<o:p></o:p>

1442

1443

<span

1444

style='mso-ansi-language:EN-US'>The same applies to the �<b style='mso-bidi-font-weight:

1445

normal'>Delete markers from displacement vs. position plot�, it is just a

1446

different view of the markers.<o:p></o:p>

1447

1448

<span

1449

style='mso-ansi-language:EN-US'><o:p> </o:p>

1450

1451

<span

1452

style='mso-ansi-language:EN-US'>After cleaning up the data set and if you have

1453

started the �displacement.m� file with �[validx,validy]=displacement;�, you

1454

will find validx and validy as variables in the workspace. Right-click on them

1455

and save the selection with a different name than validx.mat and validy.mat and

1456

nect time you open up �displacement.m� choose them. <o:p></o:p>

1457

1458

<span

1459

style='mso-ansi-language:EN-US'><o:p> </o:p>

1460

1461

<span

1462

style='mso-ansi-language:EN-US'>After cleaning up and saving the data you will

1463

want to measure strain. Here also two ways you can choose. The first one is

1464

straight forward by just clicking on either �<b style='mso-bidi-font-weight:

1465

normal'>Strain Measurement between 2 Points� which will let you choose two

1466

points or �1D Average Strain Measurement�

1467

which will use all points available. The second one is to select markers with �<b

1468

style='mso-bidi-font-weight:normal'>Select Markers to Analyze� from a

1469

certain location and then jump back to the �displacement.m� and then calculate

1470

the strain. Again it is important to keep track of all �<b style='mso-bidi-font-weight:

1471

normal'>Rotate Orientation� operations since you will analyze in both cases

1472

the x-displacement versus the x-position. In case the data was not rotated, the

1473

strain in horizontal direction in the image will be measured. <o:p></o:p>

1474

1475

<span

1476

style='mso-ansi-language:EN-US'><o:p> </o:p>

1477

1478

<span

1479

style='mso-ansi-language:EN-US'>After clicking on �<b style='mso-bidi-font-weight:

1480

normal'>Strain Measurement between 2 Points� you will have to choose two

1481

points. The function will find the closest two points to where you clicked and

1482

plot the strain versus image number.<o:p></o:p>

1483

1484

<span

1485

style='mso-ansi-language:EN-US'>If you are not happy with the result you can

1486

still choose two other markers by clicking �Yes� in the next dialog. If you

1487

stick with these markers, you can either save the result as a text file (�image_1Dstrain.txt�)

1488

or just go back to the �displacement.m� dialog. Make sure you change the filename

1489

in the explorer before you run this code another time since it will just

1490

overwrite it.<o:p></o:p>

1491

1492

<span style='mso-ansi-language:

1493

EN-US'><o:p> </o:p>

1494

1495

<span

1496

style='mso-ansi-language:EN-US'><o:p> </o:p>

1497

1498

<span

1499

style='mso-ansi-language:EN-US'><!--[if gte vml 1]><v:shape id="_x0000_i1036"

1500

type="#_x0000_t75" style='width:331.5pt;height:302.25pt'>

1501

<v:imagedata src="Correlation_Guide_2_files/image023.jpg" o:title="displacement_3"

1502

cropright="11695f"/>

1503

</v:shape><![endif]--><![if !vml]><img border=0 width=442 height=403

1504

src="Correlation_Guide_2_files/image024.jpg" v:shapes="_x0000_i1036"><![endif]><o:p></o:p>

1505

1506

<span

1507

style='mso-ansi-language:EN-US'><o:p> </o:p>

1508

1509

<span style='mso-ansi-language:

1510

EN-US'>Fig. 11: From this window two points can be picked which will be used to

1511

measure the strain<o:p></o:p>

1512

1513

<span

1514

style='mso-ansi-language:EN-US'><o:p> </o:p>

1515

1516

<span

1517

style='mso-ansi-language:EN-US'><!--[if gte vml 1]><v:shape id="_x0000_i1037"

1518

type="#_x0000_t75" style='width:331.5pt;height:296.25pt'>

1519

<v:imagedata src="Correlation_Guide_2_files/image025.jpg" o:title="displacement_4"

1520

cropbottom="1147f" cropright="11695f"/>

1521

</v:shape><![endif]--><![if !vml]><img border=0 width=442 height=395

1522

src="Correlation_Guide_2_files/image026.jpg" v:shapes="_x0000_i1037"><![endif]><o:p></o:p>

1523

1524

<span

1525

style='mso-ansi-language:EN-US'><o:p> </o:p>

1526

1527

<span style='mso-ansi-language:

1528

EN-US'>Fig. 12: Strain versus image number<o:p></o:p>

1529

1530

<span style='mso-ansi-language:

1531

EN-US'><o:p> </o:p>

1532

1533

<span

1534

style='mso-ansi-language:EN-US'>After clicking on the �<b style='mso-bidi-font-weight:

1535

normal'>1D Average Strain Measurement� button, the x-displacement versus

1536

x-direction will be plotted for each image and then fitted by a linear

1537

function. The slope is the true strain which will be plotted versus the image

1538

number after all images are processed. If you choose to save the strain versus

1539

image number you will be asked where you want to save the data as an ASCII file

1540

which can be opened with matlab, excel or just the notepad.<o:p></o:p>

1541

1542

<span style='mso-ansi-language:

1543

EN-US'><o:p> </o:p>

1544

1545

<span style='mso-ansi-language:

1546

EN-US'><o:p> </o:p>

1547

1548

<span

1549

style='mso-ansi-language:EN-US'><!--[if gte vml 1]><v:shape id="_x0000_i1038"

1550

type="#_x0000_t75" style='width:342pt;height:306pt'>

1551

<v:imagedata src="Correlation_Guide_2_files/image027.jpg" o:title="displacement_5"

1552

cropbottom="1147f" cropright="11624f"/>

1553

</v:shape><![endif]--><![if !vml]><img border=0 width=456 height=408

1554

src="Correlation_Guide_2_files/image028.jpg" v:shapes="_x0000_i1038"><![endif]><o:p></o:p>

1555

1556

<span style='mso-ansi-language:

1557

EN-US'><o:p> </o:p>

1558

1559

<span style='mso-ansi-language:

1560

EN-US'>Fig. 13: The slope of the linear fit of the displacement versus position

1561

allows to plot the true strain versus image number.<o:p></o:p>

1562

1563

<span style='mso-ansi-language:

1564

EN-US'><o:p> </o:p>

1565

1566

<span

1567

style='mso-ansi-language:EN-US'><o:p> </o:p>

1568

1569

<span

1570

style='mso-ansi-language:EN-US'>If you want to analyze a special part of your

1571

sample it is best to use the �Select

1572

Markers to Analyze� button in the �displacement.m� menu. The menu which

1573

will pop up allows you to choose between different types of grids.<o:p></o:p>

1574

1575

<span

1576

style='mso-ansi-language:EN-US'>After you have chosen the markers you want to

1577

process, run �1D Average Strain

1578

Measurement� to get the strain from these markers. Clicking the �<b

1579

style='mso-bidi-font-weight:normal'>Rotate Orientation (exchange x and y)�

1580

button and running the strain analysis again will give you the strain in the

1581

perpendicular direction.<o:p></o:p>

1582

1583

<span style='font-size:12.0pt;font-family:"Times New Roman";mso-fareast-font-family:

1584

"Times New Roman";mso-ansi-language:EN-US;mso-fareast-language:DE;mso-bidi-language:

1585

AR-SA'>

1586

1587

1588

<span style='font-size:18.0pt;

1589

mso-ansi-language:EN-US'>5. Digital Image Tracking Quick Guide <o:p></o:p>

1590

1591

<span style='mso-ansi-language:

1592

EN-US'><o:p> </o:p>

1593

1594

<span style='mso-ansi-language:

1595

EN-US'>This guide should help you to perform a simple and fast analysis of your

1596

images. Before we start you should check your image format and the naming of

1597

your files. The preferred image format is *.tif and can be compressed with the

1598

packbits compression. JPEG or other image formats as well as MPEG video

1599

compression formats will not provide you with sub pixel resolution since the

1600

images are processed to save as much space as possible. The script we use to

1601

create a list of images to process (filelist_generator.m) is kind of limited to

1602

a certain format but it is possible to generate your own list of images which

1603

will be explained later. If you want to change the format or the names of your

1604

images you can use free programs like Irfanview (<a

1605

href="http://www.irfanview.com/">www.irfanview.com</a>) to batch process a huge

1606

number of images.<o:p></o:p>

1607

1608

<span style='mso-ansi-language:

1609

EN-US'><o:p> </o:p>

1610

1611

<i style='mso-bidi-font-style:

1612

normal'>4 Steps to

1613

Success:<o:p></o:p>

1614

1615

<span style='mso-ansi-language:

1616

EN-US'><o:p> </o:p>

1617

1618

<span

1619

style='font-size:16.0pt;mso-ansi-language:EN-US'>1. Step: Filename list

1620

generation with filelist_generator.m<o:p></o:p>

1621

1622

<span

1623

style='mso-ansi-language:EN-US'><o:p> </o:p>

1624

1625

<span

1626

style='mso-ansi-language:EN-US'>Just type �filelist_generator;� and press

1627

�ENTER� at the command line of matlab. The following window should appear:<o:p></o:p>

1628

1629

<span

1630

style='mso-ansi-language:EN-US'><o:p> </o:p>

1631

1632

<span

1633

style='mso-ansi-language:EN-US'><!--[if gte vml 1]><v:shape id="_x0000_i1039"

1634

type="#_x0000_t75" style='width:243pt;height:102.75pt'>

1635

<v:imagedata src="Correlation_Guide_2_files/image001.jpg" o:title="filelist_generator"

1636

cropbottom="43901f" cropright="27208f"/>

1637

</v:shape><![endif]--><![if !vml]><img border=0 width=324 height=137

1638

src="Correlation_Guide_2_files/image002.jpg" v:shapes="_x0000_i1039"><![endif]><o:p></o:p>

1639

1640

<span

1641

style='mso-ansi-language:EN-US'><o:p> </o:p>

1642

1643

<span style='mso-ansi-language:

1644

EN-US'>Fig. 1: Input of first and last image to create an image list with

1645

filelist_generator.m<o:p></o:p>

1646

1647

<span

1648

style='mso-ansi-language:EN-US'><o:p> </o:p>

1649

1650

<span

1651

style='mso-ansi-language:EN-US'>The numbers will be the number at the end of

1652

each filename. After depositing these numbers in the dialog the next window

1653

will ask for the first 4 letters of the filenames.<o:p></o:p>

1654

1655

<span

1656

style='mso-ansi-language:EN-US'><o:p> </o:p>

1657

1658

<span

1659

style='mso-ansi-language:EN-US'><!--[if gte vml 1]><v:shape id="_x0000_i1040"

1660

type="#_x0000_t75" style='width:243pt;height:1in'>

1661

<v:imagedata src="Correlation_Guide_2_files/image003.jpg" o:title="filelist_generator_2"

1662

cropbottom="50376f" cropright="27208f"/>

1663

</v:shape><![endif]--><![if !vml]><img border=0 width=324 height=96

1664

src="Correlation_Guide_2_files/image004.jpg" v:shapes="_x0000_i1040"><![endif]><o:p></o:p>

1665

1666

<span

1667

style='mso-ansi-language:EN-US'><o:p> </o:p>

1668

1669

<span style='mso-ansi-language:

1670

EN-US'>Fig. 2: Input for the first 4 letters in filelist_generator.m<o:p></o:p>

1671

1672

<span

1673

style='mso-ansi-language:EN-US'><o:p> </o:p>

1674

1675

<span

1676

style='mso-ansi-language:EN-US'>The next step is to save the file name list

1677

into the folder with the images to process.<o:p></o:p>

1678

1679

<span

1680

style='mso-ansi-language:EN-US'><!--[if gte vml 1]><v:shape id="_x0000_i1041"

1681

type="#_x0000_t75" style='width:252pt;height:162pt'>

1682

<v:imagedata src="Correlation_Guide_2_files/image005.jpg" o:title="filelist_generator_3"

1683

cropbottom="31426f" cropright="25789f"/>

1684

</v:shape><![endif]--><![if !vml]><img border=0 width=336 height=216

1685

src="Correlation_Guide_2_files/image006.jpg" v:shapes="_x0000_i1041"><![endif]><o:p></o:p>

1686

1687

<span

1688

style='mso-ansi-language:EN-US'><o:p> </o:p>

1689

1690

<span style='mso-ansi-language:

1691

EN-US'>Fig. 3: Dialog to save the file name list into the folder with the

1692

images to analyze.<o:p></o:p>

1693

1694

<span

1695

style='mso-ansi-language:EN-US'><o:p> </o:p>

1696

1697

<span

1698

style='mso-ansi-language:EN-US'><o:p> </o:p>

1699

1700

<p class=MsoNormal style='margin-left:.5in;text-align:justify;text-indent:-.25in;

1701

mso-list:l3 level1 lfo4;tab-stops:list .5in'><![if !supportLists]><span

1702

style='font-size:16.0pt;mso-ansi-language:EN-US'>2.<span

1703

style='font:7.0pt "Times New Roman"'>    <![endif]><span

1704

style='font-size:16.0pt;mso-ansi-language:EN-US'>Run tracking with peak_labelling.m,

1705

pickpeak.m and strain_lineprofile.m:<o:p></o:p>

1706

1707

<span

1708

style='mso-ansi-language:EN-US'><o:p> </o:p>

1709

1710

<b

1711

style='mso-bidi-font-weight:normal'>�peak_labelling.m�<o:p></o:p>

1712

1713

<span

1714

style='mso-ansi-language:EN-US'>If you trust the automatic peak labeling, you

1715

can use �peak_labelling.m�. Type �[validx,validy]=peaklabelling;� It will scan

1716

your image and subtract the hopefully dark background and identify maxima with

1717

a value higher then a certain grey value. It will ask you for an image (the

1718

base image) which will be used to identify the peaks and run a first fit

1719

through all of them.<o:p></o:p>

1720

1721

<span

1722

style='mso-ansi-language:EN-US'><o:p> </o:p>

1723

1724

<span

1725

style='mso-ansi-language:EN-US'><!--[if gte vml 1]><v:shape id="_x0000_i1042"

1726

type="#_x0000_t75" style='width:252pt;height:164.25pt'>

1727

<v:imagedata src="Correlation_Guide_2_files/image007.jpg" o:title="grid_generator_1"

1728

cropbottom="30952f" cropright="25789f"/>

1729

</v:shape><![endif]--><![if !vml]><img border=0 width=336 height=219

1730

src="Correlation_Guide_2_files/image008.jpg" v:shapes="_x0000_i1042"><![endif]><o:p></o:p>

1731

1732

<span

1733

style='mso-ansi-language:EN-US'><o:p> </o:p>

1734

1735

<span

1736

style='mso-ansi-language:EN-US'>Fig. : Open the first image for automatic peak

1737

identification.<o:p></o:p>

1738

1739

<span

1740

style='mso-ansi-language:EN-US'><o:p> </o:p>

1741

1742

<span

1743

style='mso-ansi-language:EN-US'>After opening the first image �peak_labelling�

1744

will plot the image as intensity plot where blue is low and red is high

1745

intensity. You will be asked to draw a box in which �peak_labelling� will check

1746

for peaks. After selecting the area, it will take some time to process. After

1747

identifying the peaks �peak labeling will automatically start to fit all peaks

1748

and plot the residuals of all peaks. Minimizing the matlab console window will

1749

increase processing speed. The title in the figure will indicate the status of

1750

the processing and the estimated time it will take.<o:p></o:p>

1751

1752

<span

1753

style='mso-ansi-language:EN-US'><o:p> </o:p>

1754

1755

<span

1756

style='mso-ansi-language:EN-US'><!--[if gte vml 1]><v:shape id="_x0000_i1043"

1757

type="#_x0000_t75" style='width:342pt;height:306pt'>

1758

<v:imagedata src="Correlation_Guide_2_files/image029.jpg" o:title="peaklabelling01"

1759

cropbottom="1188f" cropright="11589f"/>

1760

</v:shape><![endif]--><![if !vml]><img border=0 width=456 height=408

1761

src="Correlation_Guide_2_files/image030.jpg" v:shapes="_x0000_i1043"><![endif]><o:p></o:p>

1762

1763

<span

1764

style='mso-ansi-language:EN-US'><o:p> </o:p>

1765

1766

<span

1767

style='mso-ansi-language:EN-US'>Fig. : Select an area to find peaks.<o:p></o:p>

1768

1769

<span

1770

style='mso-ansi-language:EN-US'><o:p> </o:p>

1771

1772

<span

1773

style='mso-ansi-language:EN-US'>At the end of the processing, all relevant

1774

files (�fitxy.dat� contains the fitting parameters for each point,

1775

�validx.dat/.mat� contains the x-position of each peak, and �validy.dat/.mat�

1776

contains the y-position of each peak) will be saved in the current folder. One

1777

relevant parameter is not directly accessible since this approach is trying to

1778

automate the whole processing step but can be changed in the �peak_labelling.m�

1779

file. �It can be found in

1780

�peak_labelling.m� in line 182 and 183 where the residuals of the fits in x and

1781

y directions are validated to guarantee a flawless processing. If a fit does

1782

not work at all, the function will crash. To prevent this I let the function

1783

decide very early which peak is good or bad. Only the fits with a low residual

1784

will be used. Therefore you should make sure you do not use too high values

1785

here. You will also find a similar value in the function �sortvalidpoints.m�

1786

which is called by �peak_labelling.m� at the end of the processing to create

1787

�validx� and �validy�. You have to change this value in line 39 too, otherwise

1788

the peaks will be deleted at the end. Since the input file fitxy.dat is saved

1789

before these points are deleted, you can still play around with this value and

1790

see how it affects your resulting validx and validy.<o:p></o:p>

1791

1792

<span

1793

style='mso-ansi-language:EN-US'><o:p> </o:p>

1794

1795

<span

1796

style='mso-ansi-language:EN-US'><o:p> </o:p>

1797

1798

<b

1799

style='mso-bidi-font-weight:normal'>�pickpeak.m�<span

1800

style='mso-ansi-language:EN-US'><o:p></o:p>

1801

1802

<span

1803

style='mso-ansi-language:EN-US'>After starting the function by typing at the

1804

console �pickpeak;�, the function will ask you for the first image and then

1805

will need to know how many peaks you want to identify.<o:p></o:p>

1806

1807

<span

1808

style='mso-ansi-language:EN-US'><o:p> </o:p>

1809

1810

<span

1811

style='mso-ansi-language:EN-US'><!--[if gte vml 1]><v:shape id="_x0000_i1045"

1812

type="#_x0000_t75" style='width:242.25pt;height:1in'>

1813

<v:imagedata src="Correlation_Guide_2_files/image031.jpg" o:title="pickpeak01"

1814

cropbottom="50395f" cropright="27230f"/>

1815

</v:shape><![endif]--><![if !vml]><img border=0 width=323 height=96

1816

src="Correlation_Guide_2_files/image032.jpg" v:shapes="_x0000_i1045"><![endif]><o:p></o:p>

1817

1818

<span

1819

style='mso-ansi-language:EN-US'><o:p> </o:p>

1820

1821

<span style='mso-ansi-language:

1822

EN-US'>Fig. : How many peaks do you want to identify for tracking by

1823

�pickpeak.m�?<o:p></o:p>

1824

1825

<span style='mso-ansi-language:

1826

EN-US'><o:p> </o:p>

1827

1828

<span

1829

style='mso-ansi-language:EN-US'>After you typed in a number (e.g. 20), the

1830

function will present you the selected image and you can click boxes around the

1831

peaks you want to track. For each peak you have to define a box by clicking on

1832

the lower left and the upper right of each peak. The center of each box will be

1833

highlighted by a blue circle. It is very important that you choose a box which

1834

is wide enough for the curve fitting to get enough data points. But if you

1835

choose too big boxes you will trap several peaks in them and the residual of

1836

the fit will be high which the software will interpret as bad fit. A box size

1837

of 2-4 times of the visible peaks seems to be a good idea. Also it is better to

1838

choose round shaped peaks since this provides a better greyscale profile if you

1839

choose to use a gauss function for the fitting process. <o:p></o:p>

1840

1841

<span

1842

style='mso-ansi-language:EN-US'>After you picked all peaks, the software will

1843

fit all peaks which will be displayed in a small window and after the first

1844

image processed you will only see the actual image and with blue circles on top

1845

indicating the peaks which are still in the fitting process. Vanishing circles

1846

indicate that the peak could not be fitted any more.<span

1847

style='mso-spacerun:yes'>� The title in this window will tell you the

1848

approximated total processing time and how much percent of the images are

1849

processed. <o:p></o:p>

1850

1851

<span

1852

style='mso-ansi-language:EN-US'>After all images are processed, the data will

1853

be saved the same way as in the �peak_labelling.m� function. <o:p></o:p>

1854

1855

<span

1856

style='mso-ansi-language:EN-US'><o:p> </o:p>

1857

1858

<b

1859

style='mso-bidi-font-weight:normal'>�strain_lineprofile.m�<o:p></o:p>

1860

1861

<span

1862

style='mso-ansi-language:EN-US'>This function will track two greyscale maxima

1863

in a line profile which you can choose from an image. After opening the first

1864

image, the software will let you to choose a horizontal line at a vertical

1865

position in the image. The next dialog will ask you which integration width (in

1866

vertical, y-direct) you want to use. Default is 40 but you should keep in mind

1867

that it should be either much wider or much narrower than your markers. If you

1868

choose the same width and the markers are drifting in y direction the peaks in

1869

the greyscale profile will change which will translate as error into your

1870

strain analysis. <o:p></o:p>

1871

1872

<span

1873

style='mso-ansi-language:EN-US'>The calculated greyscale profile will then be

1874

plotted and you can choose two peaks. The first click should be located on the

1875

horizontal level of the background and the vertical position of the first peak

1876

and the second click should be placed at the horizontal level of the average

1877

peak amplitude of the two chosen peaks and the vertical position of the second

1878

peak. After the second click, the function will fit two gauss functions to the

1879

greyscale profile and plot a red fitting function on top of the data while

1880

processing all images. The peak positions will be saved in the file

1881

�raw_peak_results.dat� and the strain as strain_x.dat as well as a two column

1882

file with the image number in the first column and the strain in the second

1883

column. All files are tab delimited ASCII format and can be opened e.g. with

1884

excel. You cannot use �displacement.m� for the strain analysis since this data

1885

is only 1D with 2 points.<o:p></o:p>

1886

1887

<span

1888

style='mso-ansi-language:EN-US'><!--[if gte vml 1]><v:shape id="_x0000_i1044"

1889

type="#_x0000_t75" style='width:414.75pt;height:345pt'>

1890

<v:imagedata src="Correlation_Guide_2_files/image033.jpg" o:title="line_profile"

1891

cropbottom="1593f"/>

1892

</v:shape><![endif]--><![if !vml]><img border=0 width=553 height=460

1893

src="Correlation_Guide_2_files/image034.jpg" v:shapes="_x0000_i1044"><![endif]><o:p></o:p>

1894

1895

<span

1896

style='mso-ansi-language:EN-US'><o:p> </o:p>

1897

1898

<span

1899

style='mso-ansi-language:EN-US'>Fig. : Greyscale lineprofile, ready to pick two

1900

peaks.<o:p></o:p>

1901

1902

<span

1903

style='mso-ansi-language:EN-US'><o:p> </o:p>

1904

1905

<p class=MsoNormal style='margin-left:.5in;text-align:justify;text-indent:-.25in;

1906

mso-list:l3 level1 lfo4;tab-stops:list .5in'><![if !supportLists]><span

1907

style='font-size:16.0pt;mso-ansi-language:EN-US'>3.<span

1908

style='font:7.0pt "Times New Roman"'>    <![endif]><span

1909

style='font-size:16.0pt;mso-ansi-language:EN-US'>Run displacement.m:<o:p></o:p>

1910

1911

<span

1912

style='mso-ansi-language:EN-US'><o:p> </o:p>

1913

1914

<span

1915

style='mso-ansi-language:EN-US'>Please check step 4 in �4. Digital Image

1916

Correlation Quick Guide�.<o:p></o:p>

1917

1918

<span style='font-size:18.0pt;font-family:"Times New Roman";mso-fareast-font-family:

1919

"Times New Roman";mso-ansi-language:EN-US;mso-fareast-language:DE;mso-bidi-language:

1920

AR-SA'>

1921

1922

1923

<span style='font-size:18.0pt;

1924

mso-ansi-language:EN-US'>6. Extra scripts and information you might find useful<o:p></o:p>

1925

1926

<span style='mso-ansi-language:

1927

EN-US'><o:p> </o:p>

1928

1929

<span style='mso-ansi-language:

1930

EN-US'><o:p> </o:p>

1931

1932

<b style='mso-bidi-font-weight:

1933

normal'>�stress_strainmatch.m�<span

1934

style='mso-ansi-language:EN-US'>:<o:p></o:p>

1935

1936

<span style='mso-ansi-language:

1937

EN-US'>Matching stress and strain can become a pain if they were captured with

1938

different programs and/or computers, which can be the case if the strain is

1939

captured with a camera. This little script can read in stress and strain files

1940

(as long as they are ASCII files) and match the two together. It needs the �time_image.txt�

1941

which is created by the �filelist_generator.m�, the strain file and the stress

1942

file. You have to choose which column is stress and strain in each file. After

1943

it has loaded all the files the script will ask you for the time between

1944

starting the stress measurement and the first image file. The stress versus

1945

image plot shows you immediately if the chosen value makes sense and the file �stress_image_x.txt�

1946

will be written to the �Current Folder� on the hard disk.<o:p></o:p>

1947

1948

<span style='mso-ansi-language:

1949

EN-US'><o:p> </o:p>

1950

1951

<b style='mso-bidi-font-weight:

1952

normal'>�Markerplotting.m�<span

1953

style='mso-ansi-language:EN-US'>:<o:p></o:p>

1954

1955

<span style='mso-ansi-language:

1956

EN-US'>This script will plot the markers as small dots onto the analyzed

1957

images. You have to provide validx.m, validy.m, filenamelist and the images in the

1958

�Current Folder�. After staring the script you will be asked if you want to

1959

create a video or not. If you click �yes� a folder �Video_Markers� will be

1960

created and each frame captured as a *.jpg file.<o:p></o:p>

1961

1962

<span style='mso-ansi-language:

1963

EN-US'><o:p> </o:p>

1964

1965

<b style='mso-bidi-font-weight:

1966

normal'>�resume_automate_image.m�:<o:p></o:p>

1967

1968

<span style='mso-ansi-language:

1969

EN-US'>This function allows you to resume automate_image jobs after they were

1970

stopped by the user or out of other reasons (windows decided that it is more

1971

important to install an update than running the job�). You need

1972

�resultsimcorrx.txt� and �resultsimcorrx.txt� as well as �grid_x.dat� and

1973

�grid_y.dat� and �filenamelist� in the same folder as your images you want to

1974

continue to process. Your Current Directory should be the same. You can start

1975

the function for example by typing �[validx, validy]=resume_automate_image;� at

1976

the command line of matlab and then press ENTER. <o:p></o:p>

1977

1978

<b style='mso-bidi-font-weight:

1979

normal'><o:p> </o:p>

1980

1981

<b style='mso-bidi-font-weight:

1982

normal'>�jobskript.m�<span

1983

style='mso-ansi-language:EN-US'><o:p></o:p>

1984

1985

<span style='mso-ansi-language:

1986

EN-US'>This script allows you to create a list of automate_image jobs and

1987

executes them by the use of the automate_image function. It combines the first

1988

3 steps of the Correlation quick guide described in chapter 4. The script calls

1989

first the �filelist_generator� and let you store the created file list in the folder

1990

with the images. After that the script calls the �grid_generator� and stores

1991

�grid_x.dat� and �grid_y.dat� in the same folder. You can then store the job

1992

folder by clicking on the �Save� button and then proceed either with the job or

1993

create another one. You can also load stored jobs at the beginning. If you

1994

decide not to add any more jobs (click �No�) the script will give the folders

1995

one by one to �automate_image.m�.<o:p></o:p>

1996

1997

<b style='mso-bidi-font-weight:

1998

normal'><o:p> </o:p>

1999

2000

<b style='mso-bidi-font-weight:

2001

normal'>�RTCorrCode.m�<o:p></o:p>

2002

2003

<span style='mso-ansi-language:

2004

EN-US'>This function allows to process images and calculate x- and y-strain

2005

while you take them by dynamically generating a filenamelist and reading in the

2006

files. For real time image correlation you have to find out by try and error

2007

how many markers your computer can process per second. This depends on image

2008

file size, Corrsize (which you can change in �cpcorr.m�) and on the speed of

2009

your computer and how many other things it has to do besides the processing of

2010

images. <o:p></o:p>

2011

2012

<span style='mso-ansi-language:

2013

EN-US'>You can use a dedicated machine by getting your image files through a

2014

network connection and give back the strain as a voltage by a data acquisition

2015

board to the measurement computer. <o:p></o:p>

2016

2017

<span style='mso-ansi-language:

2018

EN-US'>You can either choose to let the computer stop the processing if the

2019

next image cannot be found or by the existence of a file �end.txt� in the same

2020

directory as the image files.<o:p></o:p>

2021

2022

<span style='mso-ansi-language:

2023

EN-US'>To start the function you can type [validx,validy]=RTCorrCode; and then

2024

press ENTER. First you will be asked if you want to stop processing if the

2025

function cannot find a consecutively numbered image or in case the function

2026

finds �end.txt� in the �Current Folder�. If you want to use RTCorrCode just to

2027

process your images you can click �Stop with image check� and the function will

2028

process all images in your folder with consecutive numbers in the image file

2029

name before the point (e.g. PIC00300.Tif). If you want to process the images on

2030

the fly you should click on �Stop with end.txt�. The function will then process

2031

all images and wait for the next one to process until a file �end.txt� exists

2032

in the folder. To stop the function just create a text file with that name in

2033

the folder. After choosing the mode, the following dialog will ask you for the

2034

first image to process which you have to select.<o:p></o:p>

2035

2036

<span style='mso-ansi-language:

2037

EN-US'><o:p> </o:p>

2038

2039

<span

2040

style='mso-ansi-language:EN-US'><!--[if gte vml 1]><v:shape id="_x0000_i1046"

2041

type="#_x0000_t75" style='width:252pt;height:164.25pt'>

2042

<v:imagedata src="Correlation_Guide_2_files/image007.jpg" o:title="grid_generator_1"

2043

cropbottom="30952f" cropright="25789f"/>

2044

</v:shape><![endif]--><![if !vml]><img border=0 width=336 height=219

2045

src="Correlation_Guide_2_files/image035.jpg" v:shapes="_x0000_i1046"><![endif]><o:p></o:p>

2046

2047

<span style='mso-ansi-language:

2048

EN-US'><o:p> </o:p>

2049

2050

<span style='mso-ansi-language:

2051

EN-US'>Fig. Open the first image for RTCorrCode.<o:p></o:p>

2052

2053

<b style='mso-bidi-font-weight:

2054

normal'><o:p> </o:p>

2055

2056

<span style='mso-ansi-language:

2057

EN-US'>The file you select here will be used to calculate all consecutive file

2058

names to process all the images in the folder. If there is file missing the

2059

code will not proceed. <o:p></o:p>

2060

2061

<span style='mso-ansi-language:

2062

EN-US'>After selecting the first image, RTCorrCode will call the �grid_generator.m�

2063

and allow you to choose a grid (see Chapt. 4 for grid_generator.m).

2064

Alternatively you can preselect a grid and hand it over to the function by

2065

starting it with �[validx,validy]=RTCorrCode(grid_x, grid_y);�. This can be

2066

handy if you always run the same samples with the same grid on it. Otherwise

2067

you can now select the grid and the function will process the first image and

2068

show a rough estimate of the frames per seconds your computer will be able to

2069

process. If you select the chosen grid, the function will now search for the

2070

next image and correlate it. After correlating the second image RTCorrCode will

2071

show a window with 4 subplots. The one in the upper left corner shows the

2072

actual image and the red crosses are the equivalent of the marker positions.

2073

The upper right diagram shows the displacement vs. marker position and the

2074

linear fit in x-direction while the lower right diagram shows the same in

2075

y-direction. Last but not least, the lower left subplot shows the x- and

2076

y-strain versus image number, where blue is x- and green is y-direction.<o:p></o:p>

2077

2078

<span style='mso-ansi-language:

2079

EN-US'>RTCorrCode will leave the following files in the Current Folder:

2080

grid_x.dat, grid_y.dat (saved by grid_generator), resultsimcorrx.txt and

2081

resultsimcorry.txt (which are saved after each processed image) and validx.dat

2082

and validy.dat (after RTCorrCode.m stopped).<o:p></o:p>

2083

2084

<span style='mso-ansi-language:

2085

EN-US'><o:p> </o:p>

2086

2087

<span style='mso-ansi-language:

2088

EN-US'><!--[if gte vml 1]><v:shape id="_x0000_i1047" type="#_x0000_t75"

2089

style='width:414.75pt;height:303pt'>

2090

<v:imagedata src="Correlation_Guide_2_files/image036.jpg" o:title="RTCorrCode"/>

2091

</v:shape><![endif]--><![if !vml]><img border=0 width=553 height=404

2092

src="Correlation_Guide_2_files/image037.jpg" v:shapes="_x0000_i1047"><![endif]><o:p></o:p>

2093

2094

<b style='mso-bidi-font-weight:

2095

normal'><o:p> </o:p>

2096

2097

<span style='mso-ansi-language:

2098

EN-US'>Fig. RTCorrCode will show a window with 4 subplots. The one in the upper

2099

left corner shows the actual image and the red crosses are the equivalent of

2100

the marker positions. The upper right diagram shows the displacement vs. marker

2101

position and the linear fit in x-direction while the lower right diagram shows

2102

the same in y-direction. Last but not least, the lower left subplot shows the

2103

x- and y-strain versus image number, where blue is x- and green is y-direction.<o:p></o:p>

2104

2105

<span style='mso-ansi-language:

2106

EN-US'><o:p> </o:p>

2107

2108

<span style='mso-ansi-language:

2109

EN-US'><o:p> </o:p>

2110

2111

<b style='mso-bidi-font-weight:

2112

normal'><o:p> </o:p>

2113

2114

<b style='mso-bidi-font-weight:

2115

normal'>Input and output files:<o:p></o:p>

2116

2117

<b style='mso-bidi-font-weight:

2118

normal'>Image files<span

2119

style='mso-ansi-language:EN-US'> should be 8 bit greyscale Tiff (*.tif) images

2120

and should be named with a increasing number at the end. <o:p></o:p>

2121

2122

<span style='mso-ansi-language:

2123

EN-US'>If you want to use �filelist_generator.m�, the filename should be

2124

something like �PIC0� or� �PIC1� plus the

2125

number at the end scaling from �0001� to�

2126

�9999�. The full name would be for the first file �PIC10001.tif�. If you

2127

need to process more than 9999 image then you have to modify

2128

�filelist_generator.m� or write an email to us. The <b style='mso-bidi-font-weight:

2129

normal'>�filenamelist.mat� is a matlab file since it was easier to combine

2130

text and numbers into one file by just saving it in this format. <o:p></o:p>

2131

2132

<b style='mso-bidi-font-weight:

2133

normal'>�time_image.txt�<span

2134

style='mso-ansi-language:EN-US'> contains the time the image was captured.

2135

Please keep in mind that using other software to change the name or the format

2136

after capturing the images can lead to a change of the date and capturing time

2137

of the images. It happens that the software will change the name of the images

2138

and the new creation date and time of each image will be the time it was

2139

renamed. Programs like Irfanview have the option to preserve the original time

2140

of the images. This option has to be checked to make sure you can match stress

2141

and strain at the end of your analysis.<o:p></o:p>

2142

2143

<b style='mso-bidi-font-weight:

2144

normal'>�grid_x.dat�<span

2145

style='mso-ansi-language:EN-US'> and �grid_y.dat�

2146

are the files containing the x- and y-pixel position of the starting grid

2147

created by the �grid_generator.m� function. If you want to create your own

2148

grids, you can do that with excel and save them as tab delimited ASCII files.

2149

Both files can be organized as column vectors or matrices, as long as they are

2150

equal. <o:p></o:p>

2151

2152

<b style='mso-bidi-font-weight:

2153

normal'>�validx.dat�<span

2154

style='mso-ansi-language:EN-US'> and �vaildy.dat�

2155

are both ASCII formatted tab delimited files which contain in columns the

2156

position of each marker for each image. <o:p></o:p>

2157

2158

<b style='mso-bidi-font-weight:

2159

normal'>�fitxy.dat�<span

2160

style='mso-ansi-language:EN-US'> will be only saved if you use

2161

�peak_labelling.m� or �pickpeak.m� and contains all fitting parameters for each

2162

peak. <o:p></o:p>

2163

2164

<span style='font-size:12.0pt;font-family:"Times New Roman";mso-fareast-font-family:

2165

"Times New Roman";mso-ansi-language:EN-US;mso-fareast-language:DE;mso-bidi-language:

2166

AR-SA'>

2167

2168

2169

<span style='font-size:18.0pt;

2170

mso-bidi-font-size:12.0pt;mso-ansi-language:EN-US'>7. Acknowledgement<span

2171

style='mso-ansi-language:EN-US'><o:p></o:p>

2172

2173

<span style='mso-ansi-language:

2174

EN-US'><o:p> </o:p>

2175

2176

<span style='mso-ansi-language:

2177

EN-US'>Prof. W. N. Sharpe J. provided some helpful hints what would be

2178

important to the user and what would be a waste of time ;-). I want to

2179

acknowledge him since it is always a pleasure to work in his lab at the JHU. <o:p></o:p>

2180

2181

<span style='mso-ansi-language:

2182

EN-US'>�<o:p></o:p>

2183

2184

</div>

2185

2186

</body>

2187

2188

</html>