结合图像灰度信息和空间信息的有意义区域分割

本文提出了一种应用图像灰度信息和空间信息的分割方法.首先利用快速的优化分水岭算法将图像分成多个小区域;其次,计算每一个小区域的特性参数,并确定各个区域之间的拓扑关系;最后用模糊C-均值聚类算法根据区域的灰度特性及空间特性进行归类,获取最终的分割结果.结果显示,该方法与阈值化模糊C-均值聚类算法相比,分割结果更加有意义,而且速度也有极大地提高.

Meaningful Region Segmentation of an Image Combined with Intensity Distribution and Spatial Information Field

This paper proposed a new scheme to segment an image into meaningful regions based on its intensity distribution and spatial information.First,an image is separated into a large number of small partitions by a highly optimized watershed algorithm;then,the characteristic parameters are calculated for each partition,and the topological relation is determined between each other;finally,fuzzy C means clustering algorithm is used to obtain the final segmentation result.The result shows that this scheme can obtain more meaningful regions with less time consumption than that by the direct fuzzy C means clustering algorithm.