摘要
提出了一种利用最大熵原则和灰度变换进行图像对比度增强的方法。在最大熵原则基础上利用条件迭代算法对图像灰度级进行最佳分类,对各分类区域进行相应的灰度变换,根据不同需要选取变换参数,在图像对比度增强同时各区域均衡性也得到很大改善。将利用条件迭代算法计算最大熵多阈值的方法与最小均方误差(LMSE)计算多阈值的方法进行比较,实验结果表明,文中所用方法在迭代次数上大大低于基于最小均方误差算法所需迭代次数,节省了图像处理时间,图像均衡化效果也相对提高。
An image contrast enhancement method based on maximum entropy principle and gray-level transformation was proposed. According to the maximum entropy principle, the image gray levels were classified optimally by iterated conditional modes, and the transformation parameters were selected to enhance the image contrast and improve region homogenization. The method of iterated conditional modes was compared with Least Mean Square Error （LMSE） method in computing multi-thresholds, The experiment result demonstrates that compared with LMSE, the proposed method spends less time and acquires greater image region homogenization result,
出处
《光电工程》
EI
CAS
CSCD
北大核心
2007年第2期84-87,119,共5页
Opto-Electronic Engineering
关键词
图像增强
最大熵原则
灰度变换
区域均衡
Image enhancement
Maximum entropy principle
Gray-level transformation
Region homogenization
作者简介
章秀华（1976-），女（汉族）．湖北天门人．博士生．主要从事光电信息检测及光电图像处理．E-mail：amyyzxh@sina.com.
