遥感影像线状地物智能化提取算法研究  

Research on Intelligent Extraction Algorithm of Linear Terrain of Remote Sensing Image

作　　者：张金梅[1] ZHANG Jinmei(Tianjin Institute of Geotechnical Investigation&Surveying,Tianjin 300191,China)

机构地区：[1]天津市勘察院,天津300191

出　　处：《测绘与空间地理信息》2020年第2期169-171,共3页Geomatics & Spatial Information Technology

摘　　要：对遥感影像线状地物的提取进行了深入研究分析,分别采用Snakes模型、二值形态学和面向对象处理等3种算法智能化提取线状地物。通过实验验证,3种智能算法都是行之有效的,且Snakes模型在各种线状地物提取上表现稳定,而二值形态学和面向对象算法在部分线状地物提取上表现较为突出,整体稳定性却不如Snakes模型。The present situation of remote sensing image feature extraction is studied in depth in this paper. The Snorkes model,mathematical morphology and object-oriented algorithm are used for intelligent extraction of linear features. The experimental results show that both intelligent extraction algorithms are effective. And the Snakes model shows stability in the extraction of various existing features. The mathematical morphology algorithm and object-oriented algorithm are slightly prominent in the extraction of some linear objects,and the overall stability is not as good as the Snakes model.

关 键 词：线状地物 面向对象 SNAKES模型 二值形态学 

分 类 号：P237[天文地球—摄影测量与遥感]