一种新型阈值分割的军事地图等高线识别与提取  

A New Threshold Segmentation of Identification and Pick-up for Contour Line in Paper Military Map

作　　者：潘伟[1] 

机构地区：[1]沈阳炮兵学院电子侦察系,辽宁沈阳110162

出　　处：《计算技术与自动化》2009年第2期85-87,114,共4页Computing Technology and Automation

基　　金：国家自然科学基金资助项目(69974007)

摘　　要：针对纸制军事地图等高线的提取问题,运用图像处理、色度分析学和模式识别等理论,分析扫描地图数据产生颜色误差的原因,并根据各颜色空间的模型特点,讨论色彩空间转换的必要性。先用最大方差自动取阈法得到地图黑色要素层的单版图,再对阈值分割后的图像,进行-均值聚类,得到等高线(棕色要素)单版图,最后得到分离的等高线图层。实验研究表明方法的可用性和有效性。Aiming at the picking- up problem of contour line of military map made of paper, the theories of image segmentation, chroma analysis and mode identification, etc are applied to analyze the cause of color error in the data of the scanned map. According to the characters of model in each color space, the necessity of color space transformation is discussed and the method of the image segmentation is presented. The results of experiment research show the effectiveness of the proposed method.

关 键 词：等高线 RGB模型 HSV模型 阈值分割 聚类 

分 类 号：TP31[自动化与计算机技术—计算机软件与理论]