Research Article
Image Splicing-Based Forgery Detection Using Discrete Wavelet Transform and Edge Weighted Local Binary Patterns
Table 3
Comparison with recent state-of-the-art image splicing detection methods.
| | Approaches | Dataset used | ACC | TPR | FPR | AUC |
| | Proposed | DVMM | 97.21 | 97.87 | 3.88 | 0.98 | | CASIA v1.0 | 98.59 | 99.34 | 1.24 | 0.99 | | CASIA v2.0 | 98.33 | 97.32 | 0.96 | 0.99 | | Combined | 98.95 | 99.91 | 2.05 | 0.99 | | [5] | DVMM | 96.39 | 97.92 | 4.46 | 0.97 | | CASIA v1.0 | 94.89 | 92.30 | 2.77 | 0.94 | | CASIA v2.0 | 97.33 | 98.50 | 3.47 | 0.97 | | [31] | DVMM | 91.14 | 93.07 | 16.14 | — | | CASIA v1.0 | 96.17 | 97.65 | 6.84 | — | | CASIA v2.0 | 97.86 | 98.82 | 2.79 | — | | [32] | Composite | 93.21 | — | — | — | | [6] | CASIA v1.0 | 90.18 | 93.00 | 2.11 | — | | CASIA v2.0 | 96.21 | 93.00 | 2.90 | — | | Composite | 94.64 | 93.00 | 7.20 | — | | [33] | CISE | 93.36 | 92.99 | 1.89 | — | | [34] | CASIA v1.0 | 94.29 | — | — | 0.93 | | CASIA v2.0 | 96.52 | — | — | 0.97 | | DVMM | 94.17 | — | — | 0.93 | | | CASIA v1.0 | 95.4 | — | — | — | | [18] | CASIA v2.0 | 95.6 | — | — | — | | | DVMM | 94.8 | — | — | — | | [19] | CASIA v1.0 | 95.4 | — | — | — | | CASIA v2.0 | 95.6 | — | — | — | | DVMM | 94.8 | — | — | — | | [35] | CASIA v2.0 | | | | | | [20] | CASIA v2.0 | 99.54 | 95 | | |
|
|
