Iteration Based Polarimetric SAR Image Classification
In this paper, an iteration method is proposed for supervised polarimetric synthetic aperture radar (SAR) image classification. In this iterative approach, the optimization of polarimetric contrast enhancement (OPCE) is employed for enlarging the distance between the mean values of two kinds of targets and the Fisher method is employed for reducing the variances of two distributions. Using the proposed approach, polarimetric SAR image can be classified only after a few iterations. For comparison, the authors also use the maximum likelihood (ML) classifier for classification, based on the complex Wishart distribution. The classification results of a NASA/JPL AIRSAR L-band image over San Francisco demonstrate the effectiveness of the proposed approach.
Jian Yang Xiaoli She Tao Xiong
Department of Electronic Engineering, Tsinghua University, Beijing 100084, China
国际会议
Progress in Electromagnetics Research Symposium 2007(2007年电磁学研究新进展学术研讨会)(PIERS 2007)
北京
英文
47-50
2007-03-26（万方平台首次上网日期，不代表论文的发表时间）