递归滤波与KNN的高光谱遥感图像分类方法
|
|
涂兵, 张晓飞, 张国云, 王锦萍, 周瑶
|
Hyperspectral image classification via recursive filtering and KNN
|
|
Bing TU, Xiaofei ZHANG, Guoyun ZHANG, Jinping WANG, Yao ZHOU
|
|
表1 Indian Pines高光谱图像不同算法分类精度(10%训练样本)
|
Tab.1 Indian Pines data set classification accuracy of different algorithms (10% of training samples)
|
|
| 指标 类别 | 训练样
本/个 | 测试样
本/个 | SVM | SRC | JSRC | EMP | EPF | IFRF | LMLL | RF-KNN | | CA | Alfalfa | 10 | 36 | 79.31
(7.32) | 61.58
(10.62) | 94.44
(8.10) | 94.44
(2.78) | 97.86
(6.68) | 99.73
(0.85) | 94.44
(0.00) | 94.44
(0.00) | | Corn-N | 143 | 1 285 | 78.49
(0.83) | 53.06
(2.96) | 93.81
(2.00) | 87.77
(2.06) | 95.95
(2.28) | 97.55
(0.99) | 94.75
(0.37) | 98.49
(0.56) | | Corn-M | 83 | 747 | 80.74
(2.78) | 50.33
(3.65) | 91.51
(1.19) | 92.74
(2.28) | 96.00
(2.56) | 98.59
(0.91) | 73.76
(0.00) | 98.69
(0.48) | | Corn | 34 | 203 | 67.72
(5.46) | 37.56
(3.80) | 91.74
(3.54) | 85.73
(5.36) | 92.21
(5.33) | 97.62
(1.29) | 98.59
(0.00) | 99.70
(0.44) | | Grass-M | 48 | 435 | 90.28
(2.20) | 83.77
(1.80) | 92.37
(3.38) | 92.00
(2.81) | 99.00
(0.88) | 98.96
(1.07) | 95.86
(0.00) | 97.33
(3.34) | | Grass-T | 23 | 707 | 89.28
(2.00) | 91.39
(1.22) | 93.55
(1.00) | 97.63
(0.67) | 95.15
(2.97) | 99.05
(0.63) | 100.00
(0.00) | 96.94
(1.83) | | Grass-P | 15 | 13 | 83.04
(13.35) | 82.00
(8.37) | 100.00
(0.00) | 94.44
(5.56) | 97.24
(6.40) | 91.43
(18.07) | 100.00
(0.00) | 100.00
(0.00) | | Hay-W | 28 | 450 | 97.47
(0.90) | 93.30
(0.92) | 99.02
(0.71) | 99.91
(0.13) | 99.99
(0.05) | 100.00
(0.00) | 100.00
(0.00) | 100.00
(0.00) | | Oats | 15 | 5 | 46.35
(9.95) | 55.00
(13.94) | 62.00
(40.25) | 100.00
(0.00) | 100.00
(0.00) | 81.82
(17.64) | 100.00
(0.00) | 100.00
(0.00) | | Soybean-N | 150 | 822 | 80.02
(2.17) | 66.17
(3.90) | 91.08
(1.93) | 88.88
(0.78) | 92.21
(4.48) | 96.84
(1.35) | 90.83
(0.00) | 99.12
(0.89) | | Soybean-M | 246 | 2 209 | 80.84
(2.11) | 70.16
(1.47) | 97.08
(1.16) | 95.38
(0.88) | 91.71
(4.46) | 98.07
(1.04) | 91.67
(0.00) | 99.29
(0.37) | | Soybean-C | 60 | 533 | 78.99
(3.24) | 45.51
(2.62) | 84.54
(4.64) | 86.83
(2.37) | 94.73
(3.60) | 98.08
(1.07) | 93.21
(0.34) | 98.16
(0.98) | | Wheat | 21 | 184 | 93.80
(2.63) | 91.74
(3.06) | 86.20
(2.89) | 99.24
(0.62) | 100.00
(0.00) | 97.60
(2.30) | 99.46
(0.00) | 99.02
(0.81) | | Woods | 127 | 1 138 | 91.96
(1.49) | 89.19
(1.51) | 99.51
(0.26) | 99.63
(0.26) | 95.11
(2.86) | 99.70
(0.35) | 97.98
(0.00) | 99.72
(0.40) | | Buildings | 35 | 351 | 73.95
(4.28) | 35.45
(1.85) | 92.08
(4.33) | 97.32
(1.69) | 95.27
(2.64) | 97.38
(1.69) | 94.87
(0.00) | 98.46
(1.47) | | Stone | 37 | 56 | 93.97
(4.13) | 89.88
(2.64) | 94.29
(4.96) | 98.57
(0.80) | 96.40
(3.11) | 95.98
(5.05) | 94.64
(0.00) | 99.64
(0.80) | | OA | — | — | 83.33
(0.77) | 68.47
(0.59) | 94.19
(0.57) | 94.41
(0.84) | 94.47
(1.80) | 98.23
(0.18) | 94.45
(0.78) | 98.82
(0.29) | | AA | — | — | 81.64
(1.30) | 68.51
(1.89) | 91.45
(1.27) | 94.41
(0.84) | 96.18
(0.89) | 96.79
(1.80) | 95.69
(0.51) | 98.65
(0.33) | | Kappa | — | — | 80.90
(0.90) | 64.02
(0.63) | 93.37
(0.65) | 92.77
(0.49) | 93.67
(2.07) | 97.98
(0.21) | 93.65
(0.90) | 98.69
(0.29) |
|
|
|