一种光照不均匀图像的自适应校正算法  

Adaptive correction algorithm for non-uniform illumination images

作　　者：王殿伟[1,2] 王晶[3] 许志杰[3] 刘颖[1,2] 

机构地区：[1]西安邮电大学通信与信息工程学院,陕西西安710121 [2]电子信息现场勘查应用技术公安部重点实验室,陕西西安710121 [3](英国)哈德斯菲尔德大学计算机与工程学院

出　　处：《系统工程与电子技术》2017年第6期1383-1390,共8页Systems Engineering and Electronics

基　　金：国家自然科学基金(61202183);公安部科技强警基础工作专项项目(2014GABJC024);陕西省自然科学基金(2015JM6350);陕西省教育厅专项科研计划(14JK1680)资助课题

摘　　要：为了提高光照不均匀条件下采集的图像的视觉效果,提出了一种基于改进二维伽马函数自适应亮度校正算法。首先基于光照反射成像模型,利用快速引导滤波算法提取出光照信息;然后构造了一种改进的二维伽马函数,并利用光照的分布特性动态地调整二维伽马函数的参数,实现对图像亮度的自适应校正处理,提高光照过暗区域图像的亮度值,降低光照过强区域图像的亮度值,得到增强后的图像。对多种场景图像的处理结果表明,本文算法可以有效地降低不均匀光照的影响,提高图像的质量。An adaptive correction algorithm for non-uniform illumination images is proposed, which can eliminate the impact of non-uniform illumination on imaging results. Firstly, the illumination component is extracted by the fast guided image filter based on the luminance-reflection imaging model. In succession, an improved two-dimentional Gamma function dynamically adjusted by the local distribution characteristics of illumination is constructed, which can increase the intensity value at the low-level illumination regions and vice versa at the high-level illumination regions in the luminance domain. Finally, the non-uniform illumination image is adjusted adaptively by the proposed algorithm. Experimental results of different imaging scenarios show that the proposed algorithm can reduce the impact caused by non-uniform illumination and enhance the quality of the images.

关 键 词：图像增强 光照不均匀校正 二维伽马函数 快速引导滤波 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]