针对光学和雷达协同处理信息挖掘的需求，为提高主被动遥感数据协同处理应用于土地利用/覆盖地类识别的能力，提出了一种改进的光学和雷达遥感数据融合识别方法。以意大利PAVIA 地区的ERS SAR 和Landsat TM 影像、江苏徐州矿区的ALOS PALSAR 和AVNIR-2 影像对，ALOS PALSAR和SPOT 影像对为信息源，利用改进的小波变换与色彩域变换算法进行处理，融合结果与传统的Brovey、GS、PCT、HSV、Wavelet 融合算法作定量比较，并采用支持向量机（SVM）算法以相同的训练区分别对融合前后的影像，及不同融合结果进行典型地物类型识别。通过融合影像定量指标评价和识别应用验证，结果表明改进的融合算法很好地保留了融合前影像的光谱和纹理信息，且使用融合后影像识别的精度不仅明显优于单独利用光学或雷达影像，而且比采用的传统融合算法的识别结果也有较大提高。

In order to improve the effection of optical and SAR data collaboratively processing for land use/ land cover identification, a modified optical and SAR data fusion method is proposed in this research. The modified wavelet transform and color domain transform method is applied on ERS SAR and Landsat TM optical dataset over PAVIA city of Italy, ALOS PALSAR and AVNIR-2 optical dataset, ALOS PALSAR and SPOT dataset over Xuzhou mining area of Jiangsu province. The processing outcomes were compared with results of traditional fusion methods such as Brovey, GS, PCT, HSV, Wavelet through entropy, correlation coefficient, average gradient, deviation index, quantitative fusion evaluation indices, etc. And SVM classifier was also selected to generate land use/ land cover types of original data and different fusion outcomes with the same train samples. The results were evaluated by quantitative indexes for fused images and overall accuracy for classification maps, and which demonstrates that the modified multi-method collaboration strategy not only observed spectral and texture information, but also improved classification accuracy than only optical or SAR data, even outcomes of the traditional image fusion algorithms.