Automatic driving control method for low-pass obstacles

Abstract

The invention discloses an automatic driving control method for a low-pass obstacle, and the method comprises the steps: carrying out the two-stage visual recognition of the image content collected in the continuous driving process of an automatic driving vehicle, and obtaining the precise information of a target obstacle; combining a pre-constructed image distance measurement engine, and obtaining accurate distance measurement result of low-pass obstacle and vehicle; and finally, by utilizing the identified target obstacle information and the accurate distance measurement result, realizing an adaptive obstacle avoidance decision taking the obstacle type as the guide. According to the invention, accurate detection of the low-pass obstacle is realized, driving safety measures can be flexibly adjusted in cooperation with the detection result of the low-pass obstacle, and the purpose of reasonably avoiding or passing through the low-pass obstacle in a targeted manner is achieved.

Description

technical field [0001] The invention relates to the field of automatic driving, in particular to an automatic driving control method for low-passage obstacles. Background technique [0002] Self-driving cars realize the perception of the environment through high-precision maps + positioning systems, radar detection systems, and high-definition cameras. The radar detection system can realize the distance measurement of obstacles, the high-definition camera collects images, and through machine vision recognition, it can be judged whether the obstacle is a person, a car or other things. Among them, the radar is suitable for detecting objects with a certain height within the sensing range, but the road conditions of the vehicle are more complicated. It is difficult to achieve correct detection. The low-pass obstacle mentioned here refers to the obstacle that occupies a small interval in the Z direction relative to the vehicle coordinate system (it can be simply understood as a ...

AI Technical Summary

Problems solved by technology

Among them, the radar is suitable for detecting objects with a certain height within the sensing range, but the road conditions of the vehicle are more complicated. It is difficult to achieve correct detection. The low-pass obstacle mentioned here refers to the obstacle that occupies a small interval in the Z direction relative to the vehicle coordinate system (it can be simply understood as a relatively low obstacle); and the installation Although the lidar on the self-driving car can detect low-pass obstacles and draw point cloud data, it is also limited by the capture angle of the target scene object, and the imaging resolution of low-pass obstacles is not high. Too high, the generated 3D point cloud data is also difficult to accurately describe the distance of low-pass obstacles

[0003] Therefore, the ranging effect of the existing technology on low-pass obstacles is not ideal

Images