基于改进霍夫变换的环形交叉口识别方法

空间分布模式识别对地图综合、地图匹配等具有重要意义。环形交叉口是道路微观环形模式的典型代表。本文以改进的霍夫变换检测矢量圆环为基础，提出一种环形交叉口的几何识别方法。该方法将环形交叉口的识别分为环路识别和支路识别两部分，首先通过圆环识别、均匀度优化及相似度优化3个子过程识别环路，然后再通过连通性判别、支路分类和组合支路补充3个步骤提取支路。选取英国某区域道路网数据进行测试，结果表明，本文方法能够有效识别道路网中的环形交叉口，且召回率和准确率均高于对比方法。

Roundabout Recognition Method Based on Improved Hough Transform in Road Networks

Spatial distribution pattern is significant to the map generalization and map matching. The roundabout is the typical representative of miniature ring-like pattern in road networks. Based on the principle of improved Hough transform to detect the circle, a geometric recognition method of roundabout is proposed in this paper. This method can be divided into two parts:circulating road recognition and branch recognition. Firstly, circulating road is identified by the circle recognition, uniformity optimization, and similarity optimization. Then the branch is identified by connectivity discrimination, branch classification and combined branch supplementation. The results of partial road data in UK show that the proposed method can effectively identify the roundabouts, and both the recall and precision are higher than the comparison method.