Table 4 Accuracy analysis of the classification performed on the compositions of pattern analysis and granulometric bands using polygons from Forest Data Bank/Cadastre and the object segmentation according to the reference map
From: Methods for separating orchards from forest using airborne LiDAR
Data (Forest Data Bank/Cadastre) | Accuracy | Kappa | Other forest vegetation | Orchards | ||
|---|---|---|---|---|---|---|
1-commission | 1-omission | 1-commission | 1-omission | |||
ANN + DiffK | 74.4% | 47.2% | 99.7% | 46% | 67.4% | 99.9% |
56.5% | 26.1% | 98.7% | 42% | 36.9% | 98.4% | |
57.8% | 15.5% | 85% | 55% | 30.4% | 67% | |
Composition | 94.6% | 89.2% | 100% | 90.3% | 89.2% | 100% |
97.9% | 95.3% | 94.1% | 99.7% | 99.9% | 97.1% | |
94.6% | 89.2% | 100% | 90.3% | 89.2% | 100% | |
Data (object segmentation) | Accuracy | Kappa | Other forest vegetation | Orchards | ||
1-commission | 1-omission | 1-commission | 1-omission | |||
ANN + DiffK | 93.3% | 58.7% | 99.3% | 44.9% | 93% | 100% |
90.2% | 43% | 60.2% | 40.2% | 92.6% | 96.6% | |
78.9% | 57.7% | 68.2% | 86.6% | 89.5% | 73.9% | |
Composition | 96.4% | 92.8% | 94.7% | 98.1% | 98.2% | 94.8% |
93.2% | 86.5% | 88.6% | 97.9% | 98.1% | 89.4% | |
89.7% | 79.4% | 81.2% | 97.7% | 98.1% | 84% | |