Table 17.
Table of comparison of optimized results (ORL database).
| Method | Images for training | Recognition rate | ||
|---|---|---|---|---|
| Best (%) | Average (%) | Worst (%) | ||
| Mendoza et al. [4] (FIS) |
8 | 97.50% | 94.69% | 91.50% |
| Sánchez et al. [38] (FA) |
8 | 100% | 100% | 100% |
| Sánchez et al. [39] (MGNNs + complexity) |
8 | 100% | 99.27% | 98.61% |
| Proposed method | 8 | 100% | 100% | 100% |
| Azami et al. [43] (CGA + PCA) |
5 | 96.5% | 95.91% | 95.37% |
| Ch'Ng et al. [3] (PCA + LDA) |
5 | 96.5% | 94.75% | 94% |
| Sánchez et al. [38] (FA) |
5 | 99% | 98.30% | 98% |
| Sánchez et al. [39] (MGNNs + complexity) |
5 | 98.43% | 97.59% | 94.55% |
| Proposed method | 5 | 99% | 98.5% | 98% |