摘要
等高线的提取对构建军事地理信息系统有十分重要的意义。针对扫描军事地形图提出了一种等高线提取算法,为了克服扫描引起的色彩混淆和假彩色,引入高斯核函数构造彩色分割特征集,用来更好的逼近等高线的颜色分布;并充分利用等高线的颜色和空间信息,使用松弛迭代的方法来分割等高线;最后对细化后的二值等高线进行粘连和断点处理。实验结果表明该算法对高质量和一般质量的军事地形图的等高线提取都具有实用价值。
Contour lines are significant for creating a military geographical information system. An effective method was proposed for extracting contour lines from scanned military topographical maps. Gauss kernel function was used to create a color feature set, which described contour lines color distribution better. In topographical maps segmentation, Contour lines were extracted using an iterative algorithm. The algorithm combined spatial and color space information to overcome color-alias and false colors. Finally broken contour lines were reconnected in thinned binary map. The experimental results show that this new method is valuable when applying to a number of well-conditioned and common-conditioned military topographic maps.
出处
《系统仿真学报》
CAS
CSCD
北大核心
2009年第7期1954-1958,共5页
Journal of System Simulation
关键词
军事地形图
等高线提取
彩色分割
色彩混淆和假彩色
military topographical map
contour line extraction
color segment
color-alias and false colors
