针对城区分类,利用颜色特征构造一个新颖的无监督的分类框架.首先,基于最近提出的极化合成孔径雷达(PolSAR)数据的四分量分解模型,计算了常用的颜色空间：YUV,RGB,HSI和CIELab,通过引入颜色熵量化的选择颜色特征；然后,联合纹理特征和扩展的散射功率熵,用自适应的均值漂移算法分割PolSAR图像；最后,根据基于G0分布的距离测度合并聚簇为较为匀质的地物类别.通过L波段AIRSAR数据和C波段Radarsat-2的PolSAR数据验证了提出算法的有效性,分类正确率表明,相比于已有的工作,提出的算法对于城区有较好的区分能力.

The color features were exploited in a novel framework for the unsupervised classification of urban areas in this paper. Firstly, based on the recent four-component decomposition model of the polarimetric synthetic aperture radar (PolSAR) data, the common color spaces, such as YUV, RGB, HSI, and CIELab were calculated. The color feature was quantitatively selected from these color spaces by introducing the color entropy. Then together with the texture feature and the extended scattering power entropy, the adaptive mean-shift algorithm was used to segment the PolSAR data into clusters. Finally, the clusters were merged according to the G0 distribution-based distance measurement. The proposed framework was verified by the experiments on one AIRSAR L-band and two Radarsat-2 C-band PolSAR data. The classification accuracy indicates that the proposed method has superior discriminative ability for urban areas compared with existing works.