基于Zernike形状矩的地图匹配算法  

Map matching algorithm based on shape descriptors using Zernike moments

作　　者：曹闻[1] 万明英 李润生[1] 余文杰[3] 

机构地区：[1]解放军信息工程大学测绘学院 [2]61512部队 [3]96301部队

出　　处：《计算机应用研究》2011年第7期2786-2788,2792,共4页Application Research of Computers

摘　　要：地图匹配算法的有效性和可靠性对于智能交通系统而言是非常重要的,而目前存在的地图匹配算法在一些复杂环境下(如道路交叉口)仍然不能提供合理的输出。采用D-S证据理论融合当前车辆位置信息和方向信息可以有效地扩大待匹配道路之间的差异,但在复杂路网下信息量的不足会降低其匹配精度。因此,为了提高道路网络中的地图匹配精度,提出了基于Zernike形状矩的地图匹配算法。新算法引入Zernike矩描述轨迹曲线的形状,进一步修正了错误结果。通过仿真和实验表明,新算法在复杂环境下具有较强的有效性和可靠性。Efficient and reliable map-matching algorithms are essential for intelligent traffic systems.The map matching algorithm based D-S evidence theory could effectively expand the credibility differences of the candidate routes through fusing the distance and direction information between vehicle and route.Whereas,the matching accuracy would decrease in complicated route network because of the lack of information.In order to improve the precision of map matching,this paper proposed a noval map matching algorithm based the shape descriptors using Zernike moments.The new algorithm employed Zernike moments to describe the shape of history positioning curves and modified those errors due to the lack of data.The results of simulation and experiment show the better efficiency and reliability of the estimates even for ambiguous environments.

关 键 词：地图匹配 形状描述子 Zernike距 证据理论 

分 类 号：V249.32[航空宇航科学与技术—飞行器设计]