利用ROI裁剪图像
选择主菜单→basic tools→subset data via ROIS,选择裁剪图像 在spatial subset via ROI parameters中,选择由矢量生成的ROI,在“Mask pixels outside of ROI?”项中选择yes,Mask Background Value为0
8.监督分类
打开裁剪后的图像,以RGB:543显示,在display中选择Overlay→Region of interest
样本选择完成后,在Option→Compute ROI Separability,选择分
类图像,查看分离性
Input File: 裁剪
ROI Name: (Jeffries-Matusita, Transformed Divergence)
水体 [Blue] 36084 points:
农田 [Green] 18122 points: (2.00000000 2.00000000) 沙漠 [Sienna] 200333 points: (1.99973178 2.00000000)
未利用土地 [Orchid] 57805 points: (1.99999494 2.00000000) 山体 [Purple] 355228 points: (1.98296692 2.00000000)
农田 [Green] 18122 points:
水体 [Blue] 36084 points: (2.00000000 2.00000000) 沙漠 [Sienna] 200333 points: (1.99954820 1.99999924)
未利用土地 [Orchid] 57805 points: (2.00000000 2.00000000) 山体 [Purple] 355228 points: (1.98744710 1.99928798)
沙漠 [Sienna] 200333 points:
水体 [Blue] 36084 points: (1.99973178 2.00000000) 农田 [Green] 18122 points: (1.99954820 1.99999924)
未利用土地 [Orchid] 57805 points: (1.98700177 2.00000000) 山体 [Purple] 355228 points: (1.84635561 1.96524858)
未利用土地 [Orchid] 57805 points:
水体 [Blue] 36084 points: (1.99999494 2.00000000) 农田 [Green] 18122 points: (2.00000000 2.00000000) 沙漠 [Sienna] 200333 points: (1.98700177 2.00000000) 山体 [Purple] 355228 points: (1.99987319 2.00000000)
山体 [Purple] 355228 points:
水体 [Blue] 36084 points: (1.98296692 2.00000000) 农田 [Green] 18122 points: (1.98744710 1.99928798) 沙漠 [Sienna] 200333 points: (1.84635561 1.96524858)
未利用土地 [Orchid] 57805 points: (1.99987319 2.00000000)
Pair Separation (least to most);
沙漠 [Sienna] 200333 points and 山体 [Purple] 355228 points - 1.84635561 水体 [Blue] 36084 points and 山体 [Purple] 355228 points - 1.98296692
沙漠 [Sienna] 200333 points and 未利用土地 [Orchid] 57805 points - 1.98700177 农田 [Green] 18122 points and 山体 [Purple] 355228 points - 1.98744710 农田 [Green] 18122 points and 沙漠 [Sienna] 200333 points - 1.99954820 水体 [Blue] 36084 points and 沙漠 [Sienna] 200333 points - 1.99973178
未利用土地 [Orchid] 57805 points and 山体 [Purple] 355228 points - 1.99987319
水体 [Blue] 36084 points and 未利用土地 [Orchid] 57805 points - 1.99999494 水体 [Blue] 36084 points and 农田 [Green] 18122 points - 2.00000000
农田 [Green] 18122 points and 未利用土地 [Orchid] 57805 points - 2.00000000
分类统计结果
分类精度评价混淆矩阵
Confusion Matrix: H:\\123456\\jiandu
Overall Accuracy = (646026/667344) 96.8055% Kappa Coefficient = 0.9486
Ground Truth (Pixels)
Class 水体 农田 沙漠 未利用土地