改进分水岭算法在医学图像分割中的应用  

Application of Medical Image Segmentation Based on Improved Watershed Segmentation Algorithm

作　　者：吴定允[1] 张利红[1] 

机构地区：[1]周口师范学院物理与电子工程系,河南周口466001

出　　处：《电视技术》2011年第5期106-108,共3页Video Engineering

摘　　要：医学图像存在病变区域和背景区域,病变区域是分割的重点,亦即感兴趣区域。针对传统的分水岭变换由于对噪声和细密纹理的敏感会产生严重的过分割现象,提出两点改进:首先对原始图像运用高频强调滤波作图像增强预处理,补偿图像轮廓、突出图像的边缘信息以使图像显得更加清晰;其次是采用修改了标记符的梯度图像进行分水岭分割,整个分割过程无须进行再进行后续的合并处理,降低了分割的复杂性。仿真实验结果表明,这种改进的方法不但使过分割现象得到了很好的抑制,而且医学图像中的病变区域被有效分割出来,同时该方法还支持任意形状,且与JPEG2000标准兼容,分割效果很好。There are pathological change region and background in the medical images. The pathological change region is the emphasis of image segmentation. In this paper, two methods have been introduced to resolve the over-segmented of watershed algorithm because of the sensitivity of noise and fine texture in the region-based segmentation process： firstly, the high-frequency emphasis filter is used in the original image for enhancement preprocessing, which can make up image contour, and also can highlight the image edge information so that the image looked more clearly; Secondly, the gradient-based watershed segmentation has been adopted by use of mended label symbol, and the entire segmentation process does not need region merging so that the complexity of segmentation is reduced. The improvement can not only suppress the over-segmented phenomena properly, but also segment the pathological change regions in medical images effectively. Moreover, the methods can be used in arbitrary shape ROI and is compatible with the standard of JPEG2000.

关 键 词：感兴趣区域 分水岭算法 医学图像 

分 类 号：TN915.04[电子电信—通信与信息系统]