SAR图像相干斑滤波性能评估模型

Measurement Model and Performance Evaluation for Speckle Filter of SAR Images

提出针对单极化合成孔径雷达(SAR)图像相干斑滤波算法性能的分层检验模型和综合评价方法。模型分相干斑抑制程度和目标微波后向散射系数保持程度两个层次,包含的指标有等效视数、信号杂波比、回波辐射度损失、均值偏差、空间分辨率损失和峰值旁瓣比偏差六项。综合评价运用了多指标决策(MCDM)理论,通过线性分级的模糊语义标度,运用有序加权平均(Ordered Weighted Averaging,OWA)算子对分层指标检验值构成的性能偏好矩阵进行集结,得到相干斑滤波器的性能综合评价结果。通过真实和模拟的SAR图像数据,有效地对增强Lee、Gamma MAP、EPOS、小波软阈值和Kuan滤波算法进行了性能检验和评估,得到与主观综合评价一致的结果。

This paper proposes a measurement model and a performance evaluation method for speckle filter of single polarimetric synthetic aperture radar images, including six criteria： equivalent number of looks （ENL）, target-to-clutter ratio, radiometric loss, bias of mean value, bias of spatial resolution, and bias of peak sidelobe ratio. Based on multicriteria decision making（MCDM） principle, by linear fuzzy linguistic quantifier, ordered weighted averaging（OWA） operator is used to aggregate the performance preference matrix induced by all measurement results, Finally the evaluation results are obtained. Computational experiments by use of simulated and real SAR images to verify and evaluate such algorithms as refined Lee, Gamma MAP, EPOS, wavelet thresholding, and Kuan filter show that the proposed model is effective and the result is in accordance with that of visual evaluation.