基于无限圆逼近曲线的道路特征点提取算法仿真研究

Simulation Research on Road Feature Point Extraction Algorithm Based on Infinite-Circle Approximation Curve

通过车载GPS(全球定位系统)测量的道路信息,通常需要将复杂的道路数据进行压缩,同时选择合适的道路特征点以获得高精度的电子地图,为车辆的组合导航系统(INS/DR/MM)提供参考依据。根据载体的轨迹段可以视为一段圆弧的特性,基于无限圆逼近曲线方法研究了道路特征点提取算法,通过数据点曲率计算曲率半径,确定轨迹圆上的步长,从而提取道路特征点。实验结果表明,所提出的算法可降低计算量,从而提高数据的压缩效率,保证道路不失真。

High -precision electronic map is usually composed of compressed road information obtained from vehicle GPS （Global Positioning System） and appropriate road feature points, which can provide ref- erence for the vehicle integrated navigation system （INS/DR/MM）. According to the characteristics of the trajectory of the carrier which can be regarded as the circular curve, road feature point extraction algorithm was studied based on the method of infinite - circle approximation curve. The curvature radius was calcu- lated by the curvature of the data point to determine the step size on the trajectory circle and extract the road features. The experimental results show that the proposed algorithm can reduce the calculation amount, improve the efficiency of data compressing, and ensure the road in the map not distorted.