基于骨架边缘提取的遥感图像清晰度评价方法  

Sharpness Assessment for Remote Sensing Image Based on Abstracting the Edge Image of Skeleton

作　　者：王俊琦[1,2] 张立国[1] 付天骄[1] 张继森[1] 

机构地区：[1]中国科学院长春光学精密机械与物理研究所,吉林长春130033 [2]中国科学院大学,北京100049

出　　处：《激光与光电子学进展》2015年第9期110-116,共7页Laser & Optoelectronics Progress

基　　金：国家863计划(863-2-5-1-13B)

摘　　要：光电遥感相机由于受到大气湍流,成像离焦,机械振动等影响,导致获取的遥感图像呈现降质模糊状态,在对遥感图像进行复原处理之后,需要评估图像的复原质量。提出了一种清晰度评价方法,通过对图像进行骨架边缘提取,计算出骨架边缘对应的像素数目,并以此作为图像的清晰度参数,将降质图像与复原图像的清晰度参数作对比,得出复原后图像的清晰度提升率。通过对不同遥感图像的实验,验证了使用该方法进行图像清晰度评估的有效性与可靠性,同时证实了该方法具有抗噪声性能以及对弱振铃波纹的抑制效果。The remote sensing images obtained by optical remote sensing camera are degraded and blurred due to atmospheric turbulence, imaging defocus, mechanical vibration. It is necessary to assess the recovery quality of image after the processing of remote sensing image recovery. An assessment method of sharpness is proposed, through the study of the skeleton of image edge detection, the number of pixel on the edge of abstracting the skeleton image is calculated. And the data is used as sharpness parameter of the images. Comparing the sharpness parameter of the recovery images with the distorted images, it is sure that the sharpness of the recovery images is improved. The experiment on remote sensing images demonstrates the validity and credibility of the proposed method, in the same time, it has better results in the noise resistance and the inhibiting effect of the weak ringing ripples.

关 键 词：遥感 骨架边缘 图像质量评价 图像清晰度 信噪比 

分 类 号：TP751.1[自动化与计算机技术—检测技术与自动化装置]