Table 2 Classification accuracy assessment of 2007 image using error matrix
From: Using satellite data to monitor land-use land-cover change in North-eastern Latvia
Accuracy assement for supervised classification | |||||||||
|---|---|---|---|---|---|---|---|---|---|
Class types determined from reference source | User’s accuracy | ||||||||
Water | Wetland | Forest | Agriculture | Bare field | Urban/suburban | Totals | |||
Class types determined from classified map | Water | 50 | 0 | 0 | 0 | 0 | 0 | 50 | 100.00% |
Wetland | 0 | 36 | 0 | 0 | 0 | 0 | 36 | 100.00% | |
Forest | 0 | 0 | 48 | 3 | 5 | 1 | 57 | 84.21% | |
Agric | 0 | 5 | 2 | 45 | 5 | 6 | 63 | 71.43% | |
Bare field | 0 | 1 | 0 | 1 | 35 | 1 | 38 | 92.11% | |
Urban/suburban | 0 | 0 | 0 | 1 | 0 | 40 | 41 | 97.56% | |
Totals | 50 | 42 | 50 | 50 | 45 | 48 | 285 | ||
Producer’s accuracy | 100% | 86% | 96.00% | 90.00% | 77.78% | 83.33% | Total accuracy : | 89.12% | |
Overall Kappa statistics = 0.869 | |||||||||