摘要
虚线的矢量化一直是地形图自动识别的难点之一。文章通过人眼对虚线视觉感知特性的分析 ,提出了基于点与点之间相互作用的视觉组织模型。根据感知的类似性对地图图像进行形态滤波 ,然后运用有界深度优先搜索法生成初始种子段 ,提出可变形模型迭代算法实现对虚线的自动跟踪矢量化。从理论分析及实验结果得出 ,本算法是行之有效的 ,具有稳定性好、易于实现的特点。
Automatic vectorization of dotted lines in topographic maps is one of difficult issues for map interpretation and Geographic Information Systems. Based on the analysis of the visual perceptual organization of dotted lines,a model describing the mutual affect among dots is presented. Conforming to the perceptual rules,a serial algorithm is put forward. Firstly,the image is filtered by mathematical morphology according to the perceptual rules of similarity. Then the limited depth prior search is applied to generate the seed segments. At last the automatic vectorization of dotted lines is implemented with deformable template tracking. Experiments prove that the algorithm is practical and robust.
出处
《计算机应用》
CSCD
北大核心
2004年第1期63-66,共4页
journal of Computer Applications
基金
总装备部"十五"重点项目 (1 0 4 0 50 2 0 2 )
关键词
地形图
视觉组织
虚线
要素提取
矢量化
topographic maps
perceptual organization
dotted lines
elements recognition
vectorization.
