摘要
传统的SAR(合成孔径雷达)图像质量评价方法没有考虑视觉感知特性,导致主客观评价不一致。为了改善SAR图像质量评估方法的感知特性,提出了一种基于纹理的符合人类视觉系统感知性的SAR图像质量评估方法。由于SAR图像具有很强的纹理性,所以提取SAR图像的灰度共生矩阵并计算它的对比度,然后利用视觉对比敏感度函数与小波变换的倍频程特性来加权纹理特征的结构信息。仿真实验结果表明:具有感知特性的SAR图像质量评估方法较之传统方法更充分的利用了纹理结构特性,不仅能够反映出干扰对不同纹理区域的影响,而且能与人类视觉系统保持高度一致。通过对SAR图像进行感知质量评估可以为SAR的干扰效能提供更有效的依据,有利于干扰技术和干扰样式的优化。
Without considering perception characteristics of the human visual system （ HVS）, traditional SAR （ synthetic aperture radar） image quality assessment methods often lead to inconsistency in subjective and objective eval- uations. A texture-based SAR image quality assessment algorithm that accords with human visual perception is pro- posed in this paper. A SAR image has a strong texture property, thus the gray level co-occurrence matrix of this image is extracted and its contrast is calculated; then weigh the structural information of texture characteristics by combining the characteristics of the visual contrast sensitivity function （CSF） with frequency of wavelet transform. The results of simulation indicate that the proposed method makes better use of the texture and structure features, and it can not only assess the jamming effect more accurately in different texture regions, but also keep highly consistent with the human visual system. The proposed method affords better warranty to evaluate the interference effect, which is advantageous to the interference technology and interference pattern optimization.
出处
《哈尔滨工程大学学报》
EI
CAS
CSCD
北大核心
2015年第8期1137-1142,共6页
Journal of Harbin Engineering University
基金
国家自然科学基金资助项目(61201238)
关键词
SAR图像
纹理
人类视觉系统
图像质量评估
感知特性
灰度共生矩阵
对比敏感度函数
SAR image
textural features
image quality assessment
perception characteristics
gray level co-occurrence matrix
contrast sensitivity function
