一种基于边缘密度与自匹配系数的匹配区选取准则

景象匹配区的选取工作是影响景象匹配系统性能的重要因素.在选取准则研究方面按照选取参数的不同,目前有基于图像方差的选取准则、基于图像独立像元数的选取准则、基于图像重复模式的选取准则、基于自匹配系数的选取准则等.首先对这些准则进行了基本介绍,然后,通过具体的选取实验,指出它们在基准图选取时存在的问题,最后,针对这些问题,提出一种基于图像边缘密度与自匹配系数的景象匹配区选取准则加以解决.通过对典型地形、地貌特征的图像进行选取实验,证明了算法的有效性.

On a selection criterion of scene matching field based on edge density and self-matching coefficient

The selection of the scene matching field is an important factor which affects the optical scene matching system.There are different kinds of selection criteria of scene matching field based on different parameter,such as the selection criteria based respectively on image variance,independent pixel number,image repeating mode,self-matching coefficient and so on.In this paper,these selection criteria are introduced briefly.The problems existed are presented through the selection experiments.In order to solve the problems,a new selection criterion based on edge density and self-matching coefficient is presented.Experiments proved that this arithmetic is more effective than the others.