基于C1C2C3彩色不变特征的阴影检测与辐射恢复研究

阴影是遥感影像的基本特征之一,它将干扰图像匹配、图象变化检测或图像中信息的提取,而去除阴影则是遥感图像处理的难题。本文利用C1C2C3彩色不变特征中的C3将阴影与黑色区域分解开来;用方差作纹理滤波来提取阴影区边界;参考非阴影区相同地物像素亮度值来恢复阴影区像素的亮度值。并且用IKONOS城市区域影像进行了实验,实验表明该方法可真实再现阴影区地物特征,增加影像信息量,提高数据质量。

Research of Shadow Detection and Radiometric Restoration Based on C1C2C3 Color Invariants

Shadow is one of the features in remote sensing image.It affects image matching,change detection or information extraction.But removing that is usually very hard in image processing.In this paper,color invariants are used to distinguish shadows from other dark objects.In order to identify shadow boundaries,texture-filter is adopted to calculate the local variance.The radiometric restoration of shadow areas can be taken as the reference for the brighten value of nonshadow areas.Experimental results of IKONOS city's images indicate that the proposed method can accurately retrieve the ground features within the shadow areas.And it can enhance the contents of IKONOS images and can improve the quality of remote sensing data.