Table 6. Comparison of image retrieval methods.
Average accuracies and standard deviations of 1, 3, 5, 7, and 9 content based image retrievals for 10-fold with OtoMatch, traditional KNN, and SVM classifications using handcrafted features.
| Retrieve # Image | Similarity Measurement | OtoMatch | KNN classification and handcrafted features | SVM classification and handcrafted features | |||
|---|---|---|---|---|---|---|---|
| Average | Standard Deviation | Average | Standard Deviation | Average | Standard Deviation | ||
| Retrieve1 | Chebychev | 81.14% | 3.94% | 63.68% | 3.68% | 59.92% | 5.19% |
| Cosine | 81.71% | 6.87% | 65.59% | 6.81% | 61.00% | 5.40% | |
| Retrieve3 | Chebychev | 81.56% | 3.93% | 64.22% | 4.35% | 60.36% | 2.25% |
| Cosine | 82.33% | 4.08% | 62.78% | 4.63% | 59.92% | 3.34% | |
| Retrieve5 | Chebychev | 81.18% | 4.05% | 61.95% | 4.11% | 58.29% | 2.87% |
| Cosine | 82.60% | 3.89% | 61.18% | 3.87% | 57.06% | 2.02% | |
| Retrieve7 | Chebychev | 81.79% | 4.13% | 60.07% | 4.31% | 57.23% | 3.45% |
| Cosine | 82.47% | 4.12% | 59.43% | 3.81% | 56.41% | 3.53% | |
| Retrieve9 | Chebychev | 81.36% | 4.42% | 58.61% | 3.72% | 55.42% | 3.54% |
| Cosine | 82.26% | 3.66% | 59.19% | 3.50% | 56.27% | 2.83% | |