基于直方图的模糊最大指数熵图像分割方法

针对最大熵阈值分割算法的计算缺陷，提出了一种基于直方图的模糊最大指数熵阈值图像分割新算法．该算法把模糊性指数、模糊熵概念应用到图像分割中，结合基于灰度图像直方图的模糊最大熵阈值图像分割理论，给出了模糊最大熵的新定义，同时引入了指数熵的概念．该算法能较好地完成图像分割，较传统分割算法具有更强的抗噪能力，为后续的图像处理提供了良好的基础．通过对真实目标灰度图像的分割和对比实验，表明本文新算法分割准确，性能优越。

Image segmentation by histogram using fuzzy maximum exponential entropy

To overcome the defects of image segmentation by maximum entropy,a new method of image segmentation was presented by using fuzzy maximum exponential entropy based on histogram. In this paper, the index of fuzziness and fuzzy entropy were applied to image segmentation, the new definition of fuzzy maximum entropy was proposed and the concept of exponential entropy was imported. This algorithm can segment an image successfully, has a higher ability to remove noise than conventional thresholding image segmentation and offers a better foundation for later image processing. The experiments were conducted on a few real object images. The results show that the proposed approach is more effective and has better performance.