毫米波雷达目标点云稳健自适应聚类方法

针对毫米波调频连续波多输入多输出(frequency modulated continuous wave multiple-input multiple-output, FMCW MIMO)雷达点云目标远疏近密的密度不一问题,提出了一种基于密度聚类的稳健自适应三维点云聚类方法。首先,从原始数据中提取目标的距离、方位角以及俯仰角信息;其次,结合雷达的距离分辨率和角度分辨率将提取得到的三维信息以体素形式进行表示,并计算每个体素相应的局部度量值;再次,根据局部度量值计算各个体素的聚类搜索区域;最后,结合遗传算法(genetic algorithm, GA)自适应寻找聚类过程中所需的最佳参数并实现聚类。实验结果表明,该方法能够实现毫米波雷达三维点云的有效聚类。

Robust and adaptive clustering for point cloud with millimeter wave radar

To solve the problem that densities of point cloud, which are obtained with millimeter wave frequency modulated continuous wave multiple-input multiple-output (FMCW MIMO) radar, vary at different distances, an approach based on density-based spatial clustering of applications with noise (DBSCAN) is proposed. The method can realize the robust and adaptive clustering for 3D point cloud. Firstly, the distance, azimuth and elevation information of the target are extracted from the original data. Then, combined with the range resolution and angle resolution of the radar, the extracted information in three directions is represented in the form of voxels, and a local value of each voxel is calculated. After that, search area of each voxel in clustering can be obtained according to the local value. Finally, Genetic Algorithm (GA) is used to find the optimal parameters of clustering. Following the general steps of DBSCAN, the clustering is to be completed. The results shows that the method can cluster 3D point cloud obtained from millimeter wave radar effectively.