基于HSV色彩空间和光谱特征的遥感图像阴影检测方法

Shadow Detection Approach for Remote Sensing Images Based on HSV Colom Space and Spectral Feature

阴影是遥感图像的重要特征之一,遥感图像阴影检测方法存在偏蓝和暗色调非阴影误检的问题。本文提出一种基于HSV色彩空间和光谱特征的遥感图像阴影检测方法,该方法依据大气散射理论,利用海岸线波段、红边波段和近红外波段构建伪彩色图像,并转换到HSV色彩空间,构造权重修正归一化饱和度—强度指数;利用海岸线波段和近红外波段构造归一化光谱比值图像;利用权重修正归一化饱和度—强度指数和归一化光谱比值图像组建联合阴影指数;最后通过Otsu对联合阴影指数阈值分割得到最终阴影检测结果。该方法对多幅WorldView-3图像进行测试,阴影检测精度达到91%以上,相比现有同类方法提高约1%。实验结果表明本文方法能够有效缓解偏蓝和暗色调非阴影误检的问题,有助于提高阴影检测精度。

Shadow is a significant property of remote sensing images.Bluish and dark non-shadow misclassification problems are still urgently to be solved in shadow detection approaches for remote sensing images.The HSV color space and spectral feature based shadow detection method was proposed for remote sensing images.The proposed method constructs the false color image with Coastal band,Red edge band and NIR band,and then converts the false color image to HSV color space,in which the weighted normalized saturation-value index(WNSVI)is constructed.The normalized spectral ratio index(NSRI)is also created with Coastal band and NIR band.Joint shadow index(JSI)is constructed by WNSVI and NSRI.Finally,the shadow detection result is obtained through binarizing the JSI image with the Otsu threshold method.The proposed method was tested with multiple WorldView-3 images,and the shadow detection accuracy reached to 91%,which is about 1%higher than the existing similar methods.Experimental results show that the proposed method can effectively alleviate bluish and dark non-shadow misclassification problems,and help improve the shadow detection accuracy.