Detecting shadows in high-resolution remote-sensing images of urban areas using spectral and spatial features

作     者：Huang, Wei Bu, Min 

作者机构：Shanghai Univ Sch Informat & Commun Engn Shanghai Peoples R China 

出 版 物：《INTERNATIONAL JOURNAL OF REMOTE SENSING》 (国际遥感杂志)

年 卷 期：2015年第36卷第24期

页      面：6224-6244页

核心收录：

学科分类：083002[工学-环境工程] 0830[工学-环境科学与工程（可授工学、理学、农学学位）] 1002[医学-临床医学] 08[工学] 09[农学] 0804[工学-仪器科学与技术] 0903[农学-农业资源与环境] 0816[工学-测绘科学与技术] 081602[工学-摄影测量与遥感] 081102[工学-检测技术与自动化装置] 0811[工学-控制科学与工程] 

基　　金：National Natural Science Youth Fund of China 

主　　题：urban area shadow region component analysis Detection algorithms shadows remote sensing images 

摘      要：In accordance with the characteristics of urban high-resolution (HR) remote-sensing images, we propose a shadow detection algorithm that combines spectral and spatial features. Rather than pixel-based shadow features, the proposed features are based on shadow regions obtained by the object-based segmentation method. First, based on the shadow ratio map, the candidate shadow pixels are acquired by the Otsu method. The candidate shadow regions can be identified using connected component analysis. In the candidate shadow regions, shadow spectral and spatial features are calculated. With these two features, the true shadow regions can be distinguished from candidate shadow regions. Experiments and comparisons indicate that our proposed algorithm is feasible and effective for shadow detection in both aerial and satellite images.