An obstacle detection method and device

Abstract

An obstacle detection method and system (100), used to improve the accuracy of obstacle recognition while avoiding a large amount of data processing, the method includes: an obstacle information processing device (110) receiving an obstacle detection data collector (120) For the description information of n obstacles collected by the target vehicle, update the obstacle information list according to the description information of the n obstacles, and obtain the updated obstacle information list; the obstacle information processing device (110) according to the updated obstacle information object information list to obtain target area information and non-target area information, and send the target area information and non-target area information to the laser radar data collector (130); the obstacle information processing device (110) receives the laser radar data collector (130) ) collection of sampling points; finally, the obstacle information processing device (110) obtains an obstacle detection result according to the sampling point set and the updated obstacle information list. Therefore, a large amount of data processing is avoided, and at the same time, the accuracy of obstacle recognition is guaranteed.

Description

technical field [0001] The present application relates to the technical field of detection, in particular to an obstacle detection method and equipment. Background technique [0002] Obstacle detection and recognition are important factors to ensure the safety of vehicles in the state of automatic driving. Only by correctly identifying obstacles such as vehicles, pedestrians, and buildings around the vehicle can the vehicle's passable area and current road conditions be judged, and then the vehicle can be driven. Propose follow-up planning and decision-making. Therefore, the obstacle detection system in the vehicle plays an important role in ensuring the driving safety of the vehicle in the state of automatic driving. [0003] In obstacle detection systems, common sensors include vision sensors, millimeter-wave radar sensors, and lidar sensors. [0004] Visual sensors, such as monocular cameras or binocular cameras, can collect video information around the vehicle, and obt...

AI Technical Summary

Problems solved by technology

However, the recognition accuracy of this method for distant obstacles is poor. Although the ability to recognize distant obstacles can be improved by improving the scanning angular resolution of the lidar sensor, a large amount of point cloud data collected from the lidar sensor Extracting the contour information of obstacles consumes a huge amount of processor resources, resulting in slower update speed and longer delay of obstacle information

Images