Target detection and tracking method and device for three-dimensional point cloud data

Abstract

The invention discloses a target detection and tracking method for three-dimensional point cloud data. The method comprises the following steps: reading original point cloud data from laser radar equipment, carrying out the segmentation and extraction of a ground plane point of the obtained original point cloud data, and obtaining a ground point Pg and a non-ground point Png; carrying out clustering processing on the non-ground points to generate respective independent target point cloud clusters; obtaining target data generated by clustering through calculation in the point cloud cluster; judging the generated target data to obtain a moving target and speed information of the target; obtaining prior estimation based on motion estimation of the constant-speed linear motion model; and performing state updating on prior estimation and actual observation of the target to obtain updated state which is target data obtained in real time currently, namely three-dimensional point cloud targetmatching. According to the invention, the adopted target tracking algorithm is combined with motion estimation and feature matching based on target information, and the accuracy of target tracking isimproved through comparison with the observed value of the current frame.

Description

technical field [0001] The invention relates to the technical field of laser radar detection, in particular to a target detection and tracking method and device for three-dimensional point cloud data. Background technique [0002] In recent years, my country's auto industry has maintained a steady development trend. Unmanned driving technology represents the latest development direction of the entire industry. The purpose of this technology is to comprehensively improve the safety, comfort and stability of car driving and meet higher requirements. Market demand. [0003] With the continuous development of the economy and society, the pace of urbanization continues to accelerate, and the traffic conditions on urban roads are becoming more and more complex. In the field of driverless driving, real-time pedestrian and vehicle detection is of great significance. For example, by detecting the driving status of pedestrians and vehicles on both sides of the road, the control decisi...

AI Technical Summary

Problems solved by technology

However, the outstanding problems of target detection and tracking based on sequence images are the large amount of image processing data, the poor real-time performance of the algorithm, and the difficulty in obtaining the depth information of the target.

Images