基于粗糙集的医学图像融合算法  

Medical Image Fusion Algorithm Based on Rough Sets

作　　者：郝朝玲[1] 宋余庆[1] 周其慧[1] 陈健美[1] 晏燕华[1] 

机构地区：[1]江苏大学计算机与通信工程,江苏镇江212013

出　　处：《计算机测量与控制》2013年第9期2532-2534,2540,共4页Computer Measurement &Control

基　　金：教育部博士点基金(20113227110010);江苏省高校自然基金(10KJB520004);江苏省软件与集成电路专项基金(2009[100])

摘　　要：针对医学图像中滤波效果较差,边缘模糊问题,利用粗糙集不可分辨关系理论,提出了一种基于粗糙集的医学图像融合算法;通过对源图像进行粗糙中值滤波,将噪声从图像中分离出来;并提出一种粗糙边缘检测算法,对CT图像进行边缘检测;最后,扫描CT边缘图,定位坐标的跟踪,采用绝对值选大的融合规则实现CT图与MR图的融合;实验结果表明,提出的算法较好的保持了医学图像的细节信息,融合质量较优且效率较高。Using rough set indiscernibility relations theory, a rough set--based medical image fusion algorithm is proposed, which is aimed to solve the problem of less effective filtering and edge blur. Firstly, Noisy is extracted from the pixel of the image and then use a me- dian filter to increase the image quality by using rough sets theory. Then, a edge detection agorithm is proposed to detect CT image using the rough sets. At last, scanning CT edge detection map, tracking and positioning coordinates to achieve the fusion of CT and MR using the ab- solute value of the selected fusion rules. The results of experiments on medical image show the proposed algorithm has better kept the details of the medical image and good fusion quality and high efficiency.

关 键 词：粗糙集 不可分辨关系 中值滤波 OTSU 边缘检测 医学图像融合 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]