高光谱遥感影像纹理特征提取的对比分析

邵文静,孙伟伟,杨刚

Comparison of Texture Feature Extraction Methods for Hyperspectral Imagery Classification

Wenjing Shao,Weiwei Sun,Gang Yang

表3 雄安分类精度 (%)

Table 3 Classification accuracy of Xiong'an data

地物类别 纹理特征提取方法 riLBP SLIC EMP DMP AP 3D-Gabor JBF GF SVM 复叶槭 97.76 98.59 99.54 96.86 97.79 91.72 95.33 96.19 89.34 柳树 93.78 97.46 99.90 95.27 98.83 79.65 63.49 61.11 75.31 房屋 97.43 98.80 98.47 96.74 96.65 95.19 96.59 97.08 92.93 桃树 94.17 95.31 97.13 86.78 88.81 72.97 84.24 84.58 68.11 国槐 95.68 97.16 98.92 93.84 95.66 87.17 90.98 92.29 85.33 白腊梅 98.86 98.75 99.53 98.63 99.17 95.66 99.38 99.84 93.98 草地 89.82 82.24 92.79 81.84 76.39 64.61 56.51 53.17 55.55 水域 99.08 98.52 99.31 97.39 97.96 95.36 99.19 99.92 94.71 稀疏林 92.00 81.31 69.93 47.83 27.17 14.23 0.00 0.00 17.10 菜地 92.97 83.87 93.95 86.91 81.05 36.37 5.28 3.15 34.02 杨树 97.44 96.48 96.78 90.22 95.04 86.68 95.38 96.57 82.87 玉米 96.20 94.49 98.64 94.97 94.23 79.03 93.22 94.08 73.82 梨树 97.79 96.82 99.01 96.18 94.06 85.24 86.26 86.64 82.99 大豆 66.38 85.25 93.46 80.59 80.74 71.50 98.39 98.12 60.38 OA 97.84 96.72 98.53 94.72 94.72 86.41 89.90 89.77 83.76 AA 93.52 93.22 95.53 88.86 87.40 75.38 76.02 75.91 71.89 KC 94.21 96.18 98.29 93.83 93.85 84.16 76.32 76.25 81.07