一种光照无关图阴影去除方法  

Illumination-Invariant Image for Shadow Removal Method

作　　者：韩东辰 尚雅层[1] 来跃深[1] HAN Dongchen;SHANG Yaceng;LAI Yueshen(School of Mechatronic Engineering,Xi’an Technological University,Xi’an 710021,China)

机构地区：[1]西安工业大学机电工程学院,西安710021

出　　处：《西安工业大学学报》2021年第6期691-695,共5页Journal of Xi’an Technological University

基　　金：陕西省重点研发计划(2020GY-160)。

摘　　要：为了消除阴影等光照问题对计算机或机器人视觉检测的干扰,文中依据光学成像的原理,将RGB图像转换到光照不变的2D对数色度空间,利用独立成分分析法寻找特定投影方向,提出了一种基于光照无关图的阴影消除方法。结果表明:该方法生成了对场景光照强度近似不变的灰度图,有效地去除了图像的阴影。To eliminate the interference of shadows and other lighting problems on computer or robot vision detection,this paper proposes a method of shadow elimination based on illumination-invariant image.According to the principle of optical imaging,the RGB image is converted to the 2D logarithmic chromaticity space with constant illumination,and the independent component analysis method is used to find the specific projection direction.The results show that this method generates a grayscale image that is approximately constant to the scene’s illumination intensity,and effectively removes the shadow of the image.

关 键 词：光照无关图 阴影去除 独立成分分析 投影方向 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]