一种改进的医学图像边缘检测算法

An improved edge detection algorithm for medical image

边缘检测是医学图像预处理中最重要的内容之一。本文系统分析Pal.King模糊边缘检测算法的不足,提出了一种适用于医学图像的新型模糊边缘检测算法。该算法首先使用遗传算法对传统Otsu算法进行优化来确定阈值参数,对待测图像所对应的模糊特征平面通过基于此阈值定义的隶属函数来提取,并且对图像进行模糊增强和平滑处理以达到更好的边缘检测效果。仿真结果表明,与几种经典的处理方法相比,本算法能够对所要处理的医学图像提取出更加真实和完整的边界信息。

Edge detection is one of the most important elements in medical image processing. This paper systemly analyze the drawback of Pal ＊ King fuzzy edge detection algorithm, an improved algorithm of medical image was presented. The algorithm is present as a scheme for optimal traditional Ostu algorithm using genetic algorithms to definite threshold parameters, a definition of membership function according to this threshold for extract the fuzzy characteristics plane is correspon ded by the treatment image, to do fuzzy enhancement and smooth processing of image up to the better edge detection effect. For the medical image which processed, compare with some classic methods, this algorithm can withdraw boundary infor mation true and integrity.