面向对象的高分辨率遥感影像阴影提取方法

针对高分辨率遥感影像阴影提取的问题,通过分析影像主成分变换和HIS变换的特征,提出了一种面向对象的高分辨率遥感影像阴影提取方法:首先,使用mean shift分割方法进行影像去噪和平滑;然后,结合主成分变换和HIS变换形成了一种阴影检测指数(SDI);最后,通过阈值分割提取阴影信息。选取两景影像进行了阴影提取实验。实验结果表明,SDI能有效地区分阴影与建筑物、水体、蓝色地物、植被等非阴影地物;另外,mean shift分割能有效地去除结果中斑点噪声的影响,提高阴影检测精度。

Object oriented method of shadow extraction from high resolution remote sensing imagery

Through the analysis of the shadow characteristics of principal component transformation and HIS transformation in high resolution remote sensing image,a new object-oriented method of shadow extraction from high resolution remote sensing image is proposed.Firstly,it is the use of mean shift segmentation method for image denoising and smoothing;secondly,a shadow detection index(SDI)is formed by combining the principal component transform and the HIS transform;thirdly,shadow information is extracted by thresholding segmentation.Shadows in two images were extracted using the proposed method.The result shows that SDI can effectively distinguish shadows from buildings,water,blue objects,vegetation and other non-shadow objects;moreover,the use of mean shift segmentation can effectively eliminate the influence of the results of speckle noise and improve the accuracy of shadow detection.