基于熵的自动聚焦图像清晰度评价函数仿真分析  

Image Entropy Autofocus Sharpness Evaluation Function Simulation and Analysis under Matlab

作　　者：潘雪娟[1] 朱尤攀[1,2] 浦恩昌[1] 罗琳[1] 曾邦泽[1] 赵德利[1] 李泽民[1] 孙爱平[1] 

机构地区：[1]昆明物理研究所,云南昆明650223 [2]微光夜视技术重点实验室,陕西西安710065

出　　处：《红外技术》2016年第10期838-844,共7页Infrared Technology

摘　　要：基于图像处理的自动聚焦方法,图像清晰度评价函数的选择很重要。图像熵函数作为清晰度评价函数在不同文献中有一些形式上的变化,且关于聚焦图像评价值对应函数极大值还是极小值的论断不统一。本文详细整理了自动聚焦系统中图像信息熵的不同形式,比较分析它们的物理意义、数学原理,采用灰度图像,重点考虑离焦造成的图像模糊失真,在Matlab平台下仿真分析,得出的结论对具体自动聚焦系统设计时熵函数数学公式的选用和修正提供参考。In autofocus system based on image processing, selecting a high performance sharpness evaluation function is very important. Image entropy function has some difference in form when used as sharpness evaluation function, and conclusions of evaluation result of focused image should be maximum or minimum is not consistent. This paper summarized the different forms of image information entropy used in autofocus system, compared and analyzed their physical meaning and mathematical principle, then used gray images distorted mainly due to defocus. Under Matlab simulation and analysis, it gained conclusions that will offer reference to the selection and modification of entropy function form when designing a specific auto focus system.

关 键 词：自动聚焦 熵函数 MATLAB 

分 类 号：TN911.73[电子电信—通信与信息系统]