【摘要】 提出了一种针对极化合成孔径雷达(SAR)图像的新的分类方法——基于独立分量分析(ICA)的非监督分类方法。该方法将ICA和基于模糊集理论的非监督分类方法结合起来。用ICA方法对原始极化SAR图像进行特征提取,并用模糊C均值(FCM)算法对提取出的独立分量图像进行分类。该算法可对极化SAR图像进行自动分类,并减少由相干斑噪声所引起的分类错误,且其收敛速度快、稳定性高。采用SIR-C/X-SAR数据的试验证明了该算法的有效性。更多还原

【Abstract】 With a combination of the independent component analysis (ICA) and the unsupervised classification method based on fuzzy set theory, a new method for unsupervised classification of terrain types and man-made objects is proposed by employing the polarimetric synthetic aperture radar (POLSAR) data. The ICA is used to extract the features of the POLSAR data, and the fuzzy c-means (FCM) clustering algorithm to classify the extracted independent component image. The simulation results from SIR-C/X-SAR (Spaceborne Imaging Radar-C and X-band Synthetic Aperture Radar) data. indicate that the algorithm has advantages of automating classification, small classification errors caused by speckle, fast convergence and high stability.更多还原