基于图像熵确定CLAHE参数算法研究  

Research on Calculating CLAHE ParametersAlgorithm Based on Image Entropy

作　　者：余燕娟 YU Yan-juan(College of Electronic Engineering,Zhangzhou Institute of Technology,Zhangzhou 363000,Fujian,China)

机构地区：[1]漳州职业技术学院电子信息学院,福建漳州363000

出　　处：《兰州文理学院学报（自然科学版）》2024年第6期44-50,共7页Journal of Lanzhou University of Arts and Science(Natural Sciences)

基　　金：福建省中青年教师教育科研项目(JAT210852);漳州职业技术学院校级课题(zzykyk24008)。

摘　　要：通过对不同参数下CLAHE算法处理结果得到的图像熵曲线和主观质量的观察来确定最佳的图像CLAHE处理参数选择.采用平均强度、图像熵和均方根对比度来对图像结果分析.大雾图像和动物图像处理结果表明CLAHE优于直方图均衡化算法(HE),信息熵分别为7.37 vs 6.96和6.69 vs 5.5;同时CLAHE算法还避免了因灰度聚集的过暗区域拉伸造成图像细节丢失的问题.不同类型图像分析表明用图像熵来确定CLAHE参数的方法能获得较好的图像处理效果,用较低对比度改变了图像.The best image CLAHE processing parameter selection was determined by observing the image entropy curve and subjective quality obtained by the CLAHE algorithm for different parameters.The mean intensity,image entropy and root mean square contrast were also used to analyze the image results.The results of fog image and animal image processing showed that CLAHE was better than the histogram equalization algorithm(HE),with the information entropy of 7.37 vs 6.96 and 6.69 vs 5.5,respectively;meanwhile,CLAHE also avoided the loss of image details due to the stretching.Image analysis of different types showed that the method of using image entropy to determine CLAHE parameters can obtain better image processing effect and change the image with lower contrast.

关 键 词：CLAHE 子块的大小 截断参数 图像熵 

分 类 号：TP317.4[自动化与计算机技术—计算机软件与理论]