摘要
将影像上普遍存在的阴影视为图像退化的一种特殊形式,以整体变分模型为基础,以影像上阴影区域亮度普遍较暗且较均匀、阴影区域和非阴影区域之间的反差普遍较大的特点为约束,导出了整体变分模型用于影像上阴影检测的基本算法。通过对若干幅实际影像的阴影检测实验表明,本文算法对灰度影像和彩色影像上阴影区域的检测是有效的。
The shadow phenomenon on the imagery is treated as a special kind of image deg- radation, then, the following two restrictions： 1） shadow region on the imagery usually has low brightness; 2） a great contrast are take into account usually exists between shadow re- gion and non-shadow region on the imagery. On the basis of these analyses and the general principle of total variation, an TV model based algorithm of shadow detection and extraction on imagery is presented. After the experiments with several real images, it shows that the total variation based shadow detection algorithm is valid both for the color images and the black and white images.
出处
《武汉大学学报（信息科学版）》
EI
CSCD
北大核心
2006年第8期663-666,共4页
Geomatics and Information Science of Wuhan University
基金
测绘遥感信息工程国家重点实验室开放研究基金资助项目((01)0101)
关键词
整体变分
阴影检测
阴影区域提取
total variation
shadow detection
shadow region extraction
