基于二维灰度直方图的模糊熵分割方法  

Fuzzy Entropy Segmentation Method Based on 2D Gray Histogram

作　　者：刘正光[1] 林雪燕[1] 车秀阁[1] 

机构地区：[1]天津大学电气与自动化工程学院,天津300072

出　　处：《天津大学学报（自然科学与工程技术版）》2004年第12期1101-1104,共4页Journal of Tianjin University：Science and Technology

基　　金：天津市自然科学基金资助项目(023601011).

摘　　要：传统的基于一维灰度直方图的模糊熵分割方法不能反应图像的空间信息,抗噪声能力差,因此提出了基于二维灰度直方图的模糊熵分割算法.此算法根据像素点灰度值和其邻域灰度均值,建立二维灰度直方图,并在对应目标和背景的像限内构造像素点对目标和背影的隶属度函数,从而去除噪声和边缘像素对图像分割的影响.最后通过求模糊熵的极值,得到二维图像分割点.实验结果证明,该方法具有很强的鲁棒性和抗噪能力,分割效果明显优于一维的方法,而且可以方便地推广到其他一维熵分割中.The traditional fuzzy entropy segmentation method is based on 1D gray histogram which can′t reflect room information of image and is affected by noise acutely.This paper introduces a 2D fuzzy entropy segmentation method for gray histogram.In this method,2D histogram is built in terms of the message of the gray values of every pixel and its neighboring region and the membership function is constructed only for the pixels of the object and the background without the affection of the noise and the edge in the image.The threshold value is got through maximizing fuzzy entropy.The experimental results prove that the proposed method has good robustness and good performance of resisting noise.It can be directly extended to other 1D entropy segmentation method.

关 键 词：阈值分割 模糊熵 二维灰度直方图 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]