基于影像匹配的自发地理信息道路精度评价与改善

Accuracy Evaluation and Improvement of Volunteer Geographic Information Based on Image Matching

针对无矢量参考数据自发地理信息道路精度难以评价的问题,提出了一种基于影像匹配的道路精度评价方法。首先提取影像与自发地理信息中的道路交叉口作为控制点,并将对应的交叉口进行匹配,以同名控制点的均方根误差作为自发地理信息的道路精度;然后以两组控制点分别构建Delaunay三角网,利用两组控制点的对应关系对每个三角网进行仿射变换,从而实现对自发地理信息道路的几何纠正,以提高其精度。最后以郑州市的Open Street Map道路数据进行试验,结果表明本文算法能够有效提高自发地理信息的道路精度。

Aiming at the question that it is difficult to evaluate the accuracy of the volunteer geographic information under no vector reference data condition, a road accuracy evaluation method based on imaging matching is proposed. Firstly, the road intersections between the image and the volunteer geographic information are extracted as control points, and the corresponding control points are matched, then, the root mean square error of the control points of the same name are taken as road accuracy of the volunteer geographic information ; Secondly, Delaunay triangulation network are respectively built by two sets of the control points, and affine translation for each triangulation network are conducted by using the corresponding relationship between these two sets, in order to achieve geometric correction of volunteer geographic information and improve its accuracy. Finally, the road data of Open Street Map in Zhengzhou city are utilized to do an experiment. The result shows the algorithm proposed in this paper is able to improve the road accuracy of the volunteer information effectively.