论文部分内容阅读
本文将交叉熵和模糊散度应用于图像分割中，提出了四种最优灰度阈值选取算法．其一是基于均匀分布假设的最小交叉熵算法，其二是利用后验概率的最大类间交叉熵算法，其三是类间最大模糊散度的改进算法，其四是最小模糊散度算法．针对图像阈值化分割的要求，在后两种算法中构造了一种新的模糊隶属度函数．本文采用均匀测度和形状测度比较各算法的性能．利用多种类型测试图像得到的分割结果，显示了所提算法的有效性和通用性．
In this paper, cross entropy and fuzzy divergence are applied to image segmentation, and four optimal grayscale threshold selection algorithms are proposed. One is the minimum cross-entropy algorithm based on the assumption of uniform distribution. The other is the maximum cross-entropy algorithm using posterior probability. The third is the improved algorithm of the maximum fuzzy divergence between classes. The other is the minimum fuzzy divergence algorithm. Aiming at the requirement of thresholded image segmentation, a new fuzzy membership function is constructed in the latter two algorithms. In this paper, the performance of each algorithm is compared with uniform measure and shape measure. The segmentation result obtained from many types of test images shows the validity and generality of the proposed algorithm.