ResearchArticle
Geospatial Analysis of Urban Expansion Using Remote Sensing Methods and Data: A Case Study of Yangtze River Delta, China
Table 2
Classification accuracy evaluation results.
| Year | Classification data | Farmland | Forest | Grassland | Water bodies | Construction land | Unused land | Total | Producer’s accuracy (%) | Verification data |
| 2000 | Farmland | 181 | 6 | 9 | 1 | 3 | 4 | 204 | 88.73 | Forest | 5 | 182 | 5 | 0 | 2 | 3 | 197 | 92.39 | Grassland | 6 | 6 | 180 | 2 | 2 | 4 | 200 | 90.00 | Water bodies | 1 | 0 | 1 | 190 | 4 | 4 | 200 | 95.00 | Construction land | 2 | 3 | 2 | 4 | 186 | 2 | 199 | 93.47 | Unused land | 5 | 3 | 3 | 3 | 3 | 183 | 200 | 91.50 | Total | 200 | 200 | 200 | 200 | 200 | 200 | 1200 | | User’s accuracy (%) | 90.50 | 91.00 | 90.00 | 95.00 | 93.00 | 91.50 | | |
| Overall accuracy = 91.83%; Kappa coefficient = 0.90 |
| 2017 | Farmland | 191 | 2 | 4 | 1 | 1 | 3 | 202 | 94.55 | Forest | 3 | 191 | 3 | 0 | 0 | 1 | 198 | 96.46 | Grassland | 2 | 3 | 189 | 0 | 1 | 2 | 197 | 95.94 | Water bodies | 0 | 0 | 0 | 195 | 3 | 2 | 200 | 97.50 | Construction land | 1 | 1 | 2 | 3 | 192 | 2 | 201 | 95.52 | Unused land | 3 | 3 | 2 | 1 | 3 | 190 | 202 | 94.06 | Total | 200 | 200 | 200 | 200 | 200 | 200 | 1200 | | User’s accuracy (%) | 95.50 | 95.50 | 94.50 | 97.50 | 96.00 | 95.00 | | |
| Overall accuracy = 95.67%; Kappa coefficient = 0.95 |
|
|