文章基本信息

标题：Colour map object separation
本地全文：下载
作者：Amândio Cordeiro ; Pedro Pina
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：2006
卷号：XXXVI Part 7
出版社：Copernicus Publications
摘要：This paper addresses the issue of colour objects separation from map images obtained by current commercial scanners, in the RGB system. The proposed methodology is determined by some tested principles, as follows: 1. No Tonal Change. Until its isolation, the colour-object maintains its former tonalities. 2. Unique transform. The pixels that do not belong to the objects under consideration are "sent" to the background (value 255, in the 3 components). 3. Colour Spatial Correlation. The decision of considering a non-object pixel depends only on its "cut value" in a component. A cut value divides the RGB colour cube in two separate volumes or Sections. 4. Spatial Integration. Different Sections combine to form Regions, as intersections of Sections, for the noise-object and Co-Regions, as unions of Sections, for the isolation of the colour-objects
关键词：RGB System; Colour Segmentation; Image Analysis; Map Objects Recognition; Mathematical Morphology