基于加权割的图像分割

Weighted Cut Based Image Segmentation

提出了一个新的图分割模型——加权割模型,设计了一个基于加权割的图像分割算法(Image segmentation Algorithm Based on Weighted Cut,简记为ISAWC).加权割模型的特点是:(1)整合了图像的局部和整体分割信息;(2)在加权意义下最小化加权割能同时达到类间最大相异性和类内最大一致性.本文证明可通过求解一个特征向量问题来优化加权割.模拟点集和实际图像上的实验验证了ISAWC的有效性.

A novel graph partitioning criterion, weighted cut, is presented, and its application to the image segmentation problem is demonstrated. An important characteristic of the criterion is that in the course of image segmentation the local and global image segmentation information is fused together. Moreover, optimizing weighted cut can ensure that the inter-cluster similarity is minimized while intra-cluster similarity is maximized. We show that an efficient computational technique based on an eigenvector problem can be used to optimize this criterion. The experimental results on a number of artificial point sets and real-world images show the effectiveness of the new criterion.