文章基本信息

标题：CARTOGRAPHIC DATA EXTRACTION FROM AIRBORNE IMAGERY BY HIERARCHICAL-BASED MORPHOLOGIC IMAGE PROCESSING
本地全文：下载
作者：F. Soares ; A. Navarro ; T. Santos 等
期刊名称：ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
印刷版ISSN：2194-9042
电子版ISSN：2194-9050
出版年度：2010
卷号：XXXVIII - 4/C7
出版社：Copernicus Publications
摘要：This paper presents a comprehensive approach to effectively extract cartographic urban data from high resolution satellite imagery. It consists of a sequence of image processing techniques, for image segmentation, based on RGB band separation, analysis and pre- processing, followed by a morphological-based approach for data segmentation. The chosen image objects for this study are roof-tile buildings and road network. The first step of this approach consists of a spectral response evaluation for roof-tile building objects in a dense urban environment, being those enhanced through proper sequence of standard arithmetic operators, applied to RGB bands, segmented and generalized. The second step aims at urban road network segmentation for cartographic representation purposes, by combining watershed and dual reconstruction morphological transformations, which characterize the hierarchical waterfall concept. For the latter concept, a new approach is developed in order to improve hierarchical segmentation procedure, to better induce object discrimination. Each one of the referred objects will be segmented in separate. The road network segmentation will have in consideration the previous result of roof-tile buildings extraction. Finally, segmented objects will be compared with other extraction results in order to do proper validation. The method is applied over high-resolution ortho-rectified images, taken from satellite, of the city of Lisbon. Results will show the practicality of the method for purposes of cartographic data structures acquisition, from high- resolution satellite imagery, aiming urban management and GIS applications
关键词：Hierarchical Segmentation; Urban Feature Extraction; Watershed segmentation; Waterfall segmentation