基于分水岭和种子生长的彩色图像分割

Color image segmentation algorithm based on watershed segmentation and seed growing segmentation

提出一种基于分水岭和种子区域生长的彩色图像改进分割算法。该算法先对彩色图像进行分割前预处理,使用分水岭算法对图像进行初始分割,然后根据一定的规则从分水岭算法分割形成的区域中自动选取种子区域并进行生长,最后合并相似区域或小区域。以区域作为种子生长单位,使用种子区域一次性生长方法进行生长。实验结果表明,该算法分割结果较好,分割速度较快。

A color image segmentation algorithm based on watershed segmentation and seed region growing segmentation is proposed in this paper.Firstly,the color image is preprocessed before segmenting.Secondly,watershed segmentation is used to initialize the image.Thirdly, region seeds are selected automatically to grow according to two rules from all regions produced by watershed segmentation.Finally,similar or small regions are merged by region-merging method.In this new algorithm,regions are considered as growth units with a disposable seed region growing method.The experimental results show that the algorithm can quickly produce good segmentation results.