采用模糊形态学和形态学分水岭算法的图像分割  

Image Segmentation Based on Fuzzy Mathematical Morphology and Watershed Algorithm

作　　者：陈婷婷[1] 程小平[1] 

机构地区：[1]西南大学计算机与信息科学学院,重庆400715

出　　处：《西南大学学报（自然科学版）》2008年第3期142-145,共4页Journal of Southwest University(Natural Science Edition)

摘　　要：提出了一种基于模糊形态学和形态分水岭算法的图像分割方法,先利用模糊形态学闭开运算平滑原始图像,再通过形态分水岭变换实现图像分割.实验结果表明这种方法具有较好的分割效果.Image segmentation is one of basic problems in image processing and computer vision, and is a key step in image processing and image analysis. At present thousands of methods have been put forward to image segmentation. Some use classical methods and others use new methods. Based on the detailed analysis for current approaches, a method based on fuzzy mathematical morphology was advanced for image segmentation in this paper. First, fuzzy mathematical morphology closing and opening operators were employed to smoothen the original image. Through image smoothness, important region contours were preserved while most small regular details and noise are removed. Second, the watershed transformation was used to implement image segmentation. Experimental results have proved that this method is effective for good segmentation.

关 键 词：图像分割 模糊形态学 分水岭算法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]