3D target detection and tracking method based on camera and laser radar

Abstract

The invention discloses a 3D target detection and tracking method based on a camera and a laser radar. The method comprises the following steps: firstly, performing target detection on an image acquired by a camera to obtain a 2D detection frame and assign a number; meanwhile, carrying out feature point detection, descriptor extraction and feature point matching on the image; then, performing ground segmentation on the point cloud data, and projecting the non-ground point cloud into a corresponding image according to the external parameters of the laser radar and the camera; then, screening and clustering the point clouds according to the 2D detection frame in the image to obtain an object 3D detection frame; and finally, according to the matched feature points, performing matching association on the detection frames of the front and rear frames of images, and replacing numbers to realize 3D tracking of the object. According to the method, the advantages of the laser radar and the camera data are fully fused, the defects that the laser radar classification capability is weak and the camera lacks depth information are overcome, 3D target tracking of the object is realized, and the method can be used for unmanned 3D target detection and tracking tasks.

Description

technical field [0001] The invention relates to a method for detecting and tracking a 3D target, in particular to a method for detecting and tracking a 3D target based on a camera and laser radar. Background technique [0002] At present, the development of unmanned driving technology is of great significance to improving road traffic safety, alleviating urban traffic congestion, and improving travel efficiency. The unmanned driving system mainly includes modules such as environment perception, decision planning and control execution. The environment perception system is mainly responsible for sensing the surrounding environment of the vehicle, including detecting the type and position of obstacles and tracking obstacles, etc., which is the key to realize unmanned driving. prerequisites and foundations. [0003] Vision-based object detection methods are mainly divided into two categories, namely, two-stage methods and one-stage methods. The typical representative of the tw...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to provide a 3D target detection and tracking method based on camera and laser radar to solve the problem that the existing single sensor-based perception system cannot meet the needs of automatic driving

Images