Table 4 A comparative study between different fusion methodologies
From: Adaptive polar transform and fusion for human face image processing and evaluation
Sl. no. | Author | Techniques | Database used | Reported performance |
|---|---|---|---|---|
1. | M. K. Bhowmik et al. [19] | log-polar transformed + PCA | OTCBVS (IRIS) | 93.81% |
2. | Mohammad Hanif et al. [38] | Gabor Filter Technique | Equinox | DWT - 90.31% |
OIF - 95.84% | ||||
3. | M. K. Bhowmik et al. [39] | Daubechies wavelet transform + PCA + ICA | OTCBVS (IRIS) | PCA - 91.13% |
ICA I - 94.44% | ||||
ICA II - 89.72% | ||||
4. | M. K. Bhowmik et al. [40] | Pixel fusion + RBF | OTCBVS (IRIS) | 97.05% |
5. | M. K. Bhowmik et al. [41] | Optimum + Eigenspace projection + Multilayer Perceptron | OTCBVS (IRIS) | 93% |
6. | D. Bhattacharjee et al. [35] | Eigenspace projection + Multilayer Perceptron + Backpropagation learning | OTCBVS (IRIS) | 95.07% |
7. | M. K. Bhowmik et al. [42] | Pixel fusion + CCIPCA + SVM | OTCBVS (IRIS) | 97.28% |
8. | M. K. Bhowmik et al. [43] | Wavelet transformation + multiresolution analysis + MLP + RBF | OTCBVS (IRIS) | Feature level - 87.28% |
Decision level - 94.95% | ||||
9. | M. K. Bhowmik et al. [44] | Daubechies wavelet co-efficient + ICA + MLP | OTCBVS (IRIS) | 91.5% |
10. | M. K. Bhowmik et al. [45] | Eigenspace projection + MLP + RBF | OTCBVS (IRIS) | RBF - 96% |
MLP - 95.07% | ||||
11. | D. R. Kisku et al. [46] | Dempster-Shafer decision theory + SIFT features | ORL, IITK | ORL - 98.93% |
IITK - 96.29% | ||||
12. | R. Singh et al. [47] | Match score fusion + 2ν GSVM + Dezert Smarandache theory + SVM | Equinox | DSm match score fusion - 98.08% |
SVM - 95.05% | ||||
DST - 96.51% | ||||
2ν GSVM - 94.98% | ||||
13 | Present method | Adaptive Log Polar Registered + Wavelet Fusion + PCA + MLP | OTCBVS(IRIS) | 98.36% |
UGC-JU | 95.77% |