融合高分辨率遥感影像和POI数据的多特征潜在语义信息用于识别城市功能区

高子为,孙伟伟,程朋根,杨刚,孟祥超

Identify Urban Functional Zones Using Multi Feature Latent Semantic Fused Information of High-spatial Resolution Remote Sensing Image and POI Data

Ziwei Gao,Weiwei Sun,Penggen Cheng,Gang Yang,Xiangchao Meng

表4 ETM模型中不同特征组合得到的分类精度对比 (%)

Table 4 Comparison of classification accuracy of different feature combinations in ETM Model

研究区 Spe POI Spe+Tex Spe+Tex+SURF Spe+Tex+SURF+GIST 本文方法 1 OA 37.02 62.86 44.70 64.76 76.70 85.67 Kappa 24.66 53.94 32.39 55.79 69.75 81.68 2 OA 39.26 67.13 49.78 66.96 76.90 85.78 Kappa 26.72 58.04 38.31 57.38 70.24 80.03