自动图像阈值分割算法

该文提出了一种新的图像阈值分割算法。该算法通过求取最大模糊熵准则下,灰度均值直方图的最佳模糊划分参数来确定两个模糊集A和B,图像分割阈值即选取为两个模糊集的交点。该算法用Zadth的模糊熵定义适应度函数,采用改进的遗传算法寻求最佳模糊参数。该文对遗传算法的改进包括,给出了缩短染色体码长的编码方法和性能良好的改进的单点交叉算子和均匀变异算子。实验结果表明,该算法的分割效果与二维模糊熵算法接近,而计算时间还没有用到二维模糊熵算法的一半。

A Automatic Thresholding Algorithm for Image Segmentation

A novel thresholding algorithm for image segmentation is proposed in this paper.This algorithm defines two fuzzy sets A and B on a local average histogram of the image.The threshold is selected as the crossover point of the two fuzzy sets.The paper uses genetic algorithm with special encoding method and modified crossover and mutation ge-netic operators to find the best partition parameters of A and B,and Zadth's fuzzy entropy is chosen to define the ob-jective function.Experiments show that the proposed algorithm can give the same good results with less than half com-putational time comparing with two-dimensional fuzzy entropy algorithms .