Skip to main content

Table 1 Misclassification error rate for randomly tampered images

From: Automated approach for splicing detection using first digit probability distribution features

Q1–Q2

Proposed

[7]

[10]

[12]

[11]

60–65

[11]

60–70

[11]

60–75

[11]

60–80

[11]

60–85

[11]

65–85

[11]

70–85

[11]

75–85

[11]

80–85

60–65

0.3468

0.5704

0.6099

0.1949

0.1904

0.7885

0.8291

0.8201

0.8309

0.8221

0.8220

0.8237

0.6717

60–70

0.1304

0.2774

0.2940

0.1296

0.6525

0.1116

0.8382

0.8282

0.8247

0.8375

0.8373

0.8363

0.6476

60–75

0.2190

0.0948

0.2114

0.1525

0.8168

0.8318

0.1142

0.3255

0.1546

0.2504

0.3140

0.2278

0.3332

60–80

0.1944

0.2003

0.1707

0.1495

0.6763

0.8385

0.4128

0.0859

0.7329

0.1065

0.1006

0.1445

0.8089

60–85

0.2238

0.2575

0.1675

0.1310

0.8235

0.2221

0.1699

0.8285

0.0895

0.6289

0.8207

0.5750

0.3769

65–85

0.2166

0.2743

0.1799

0.1219

0.7148

0.8372

0.3762

0.1683

0.2571

0.0980

0.2744

0.1954

0.6036

70–85

0.2205

0.1939

0.1686

0.1872

0.6201

0.8451

0.4183

0.1326

0.7141

0.2305

0.0950

0.3201

0.8004

75–85

0.2965

0.2331

0.7668

0.1693

0.7439

0.8377

0.3582

0.2831

0.4030

0.2530

0.3533

0.1272

0.8053

80–85

0.2254

0.4861

0.8517

0.2719

0.4786

0.7091

0.7978

0.8433

0.8287

0.8325

0.8449

0.8392

0.1788

  1. Q 1 - Primary compression quality of tampered image, Q 2 - Secondary compression quality of tampered image