Roadside-laser-radar-based vehicle queuing length real-time estimation method and system

The invention discloses a roadside-laser-radar-based vehicle queuing length real-time estimation method and system. The method comprises the steps: acquiring all position points scanned by a roadsidelaser radar for vehicles on a road, and obtaining to-be-processed three-dimensional point cloud data; sequentially performing background filtering on the to-be-processed three-dimensional point clouddata; performing clustering processing on the three-dimensional point cloud data after background filtering; performing target identification on the clustered three-dimensional point cloud data, and identifying vehicles on the road; performing lane recognition based on a result after target recognition; estimating the speed of each vehicle on the road based on the three-dimensional point cloud data of the adjacent frames of the laser radar; and on the basis of the speed of each vehicle on the road, determining the tail vehicle on each lane, and further estimating the queuing length on each lane. The vehicle queuing len