摘要
为有效去除城市遥感影像中阴影效应所产生的影响,实现阴影区域的自动检测及区域内景物重建,利用纹理分析的方法,通过计算共生矩阵及其相关指数来确定阴影区域,以实现阴影区域的自动检测;同时利用分形插值中的随机中点替代法对阴影区域内景物进行了复原,真实再现了阴影区域内的景物特征.对实际航空遥感影像的处理结果表明,该方法能有效实现遥感影像阴影区域的自动检测及区域内的景物重建.
To effectively remove the shadows in city＇s remote sensing image, the method based on texture analysis was used to detect the shadow region automatically and rebuild the objects in it further. Based on calculating the cooccurrence matrix and its relevant parameters, the shadow region was conformed and detected automatically. Simultaneously, the objects in shadow region were rebuilt based on the random midpoint replacement method of fractal interpolation, with the characteristic of the real objects in it retrieved. The processing result of real aerial remote sensing image proves that the shadow region can be detected automatically and the objects in it can be rebuilt with this method.
出处
《天津大学学报》
EI
CAS
CSCD
北大核心
2008年第7期800-804,共5页
Journal of Tianjin University(Science and Technology)
基金
天津市自然科学基金资助项目(07JCYBJC13800)
教育部博士点基金新教师资助项目(20070056104)
关键词
阴影检测
分形
共生矩阵
纹理分析
随机中点替代法
shadow detection
fractal
co-occurrence matrix
texture analysis
random midpoint replacement method
