Table 12.
Comparison of classification accuracies on the ICL dataset (220 species) and its two subsets (50 species each)
| Descriptor | Feature | Classifier | Accuracy | Studies |
|---|---|---|---|---|
| Full dataset: 220 species | ||||
| FD | Shape | 1-NN | 60.08 | [135] |
| TAR | Shape | 1-NN | 78.25 | [135] |
| IDCS | Shape | 1-NN | 81.39 | [135] |
| IDSC | Shape | k-nn | 83.79 | [139] |
| GF | Texture | Fuzzy k-NN | 84.60 | [136] |
| MARCH | Shape | 1-NN | 86.03 | [135] |
| HoCS | Shape | Fuzzy k-NN | 86.27 | [136] |
| MDM | Shape | Fuzzy k-NN | 88.24 | [136] |
| IDSC | Shape | Fuzzy k-NN | 90.75 | [136] |
| SIFT, SC | Shape + vein | k-NN | 91.30 | [139] |
| EnS and CDS | Shape + texture | SVM | 95.87 | [140] |
| DS-LBP | Shape + texture | Fuzzy k-NN | 98.00 | [136] |
| Subsets: 50 species | ||||
| IDSC | Shape | SVM | 95.79 (63.99) | [121] |
| FD | Shape | 1-NN | 96.00 (80.88) | [62] |
| HOG | Shape | SVM | 96.63 (83.35) | [121] |
| LBP | Shape | SVM | 97.70 (92.80) | [121] |
| IDSC | Shape | 1-NN | 98.00 (66.64) | [62, 145] |
| MDM with ID | Shape | 1-NN | 98.20 (80.80) | [62] |
| HOG | Shape | 1-NN | 98.92 (89.40) | [145] |
| I-IDSC | Shape | 1-NN | 99.48 (88.40) | [158] |
Certain studies used two subsets of the ICL leaf dataset: subset A and subset B (in brakets). Subset A includes 50 species with shapes easily distinguishable by humans. Subset B includes 50 species with very similar but still visually distinguishable shapes