基于HSI色彩空间的资源三号影像阴影检测

摘要

Due to the lack of light radiation on certain areas of remote sensing image,inevitably there will be a shadow.Shadow means loss of image information,and shadow detection for remote sensing image is of great significance in the recognition of surface features and image matching.This paper mainly introduces the shadow detection Algorithm based on HIS color space.According to the characteristics of high tone and low brightness,the shadow of remote sensing image is detected,combined with Otsu method to calculate optimal threshold of ratio image.And besides,the impact of plants and some non-shaded areas is excluded by calculating the optimal threshold for the G component in the RGB color space.In the meantime,the program is completed by using the remote sensing image of the domestic high resolution ZY-3 satellite in different seasons and different solar elevation angles of the same areas.The Experimental results show that,the method based on HIS color space shadow detection can quickly and effectively detect the shadow image,and can distinguish between trees,rivers and the other dark objects.%由于遥感影像上某些区域的光照辐射不足,不可避免地会产生阴影,阴影意味着图像信息的损失,而遥感影像的阴影检测在地物的识别和影像匹配方面具有重要意义.本文主要介绍的是基于HIS色彩空间的阴影检测方法,在检测过程中,根据阴影高色调低亮度的特性,结合大津法计算比值图像最佳阈值进行遥感影像阴影检测,并且在RGB色彩空间计算G分量的最佳阈值来排除树木植被和一些非阴影区域对阴影检测的影响.同时采用国产高分辨遥感卫星——资源三号的同一地区不同季节和不同太阳高度角的遥感数据进行阴影的对比检测.实验结果表明:本文基于HIS色彩空间的阴影检测方法可以快速有效地检测出影像上的阴影,并且能区分树木、河流等暗色物体.