Obstacle position recognition method, device, computer equipment and storage medium

Abstract

The present application relates to an obstacle position recognition method, device, computer equipment and storage medium. The method includes: acquiring depth information, plane area information, and body area information corresponding to the obstacle; determining the target obstacle when the depth information, plane area information, and body area information are similarly matched; Set the location information, calculate the distance error information between the depth information, the plane area information and the main area information and the preset location information respectively; calculate the conditional probability distribution of the sensor observation information under the preset location information according to the distance error information; The distribution is estimated to obtain the target position information corresponding to the target obstacle. Using this method can improve the accuracy of estimating the position of obstacles that partially enter the field of view of the camera, thereby improving the performance of automatic driving.

Description

technical field [0001] The present application relates to the field of computer technology, in particular to an obstacle intention recognition method, device, computer equipment and storage medium. Background technique [0002] With the development of artificial intelligence technology, self-driving technology has emerged. Self-driving cars rely on the cooperation of artificial intelligence, visual computing, radar, monitoring devices and global positioning systems, allowing computers to automatically and safely drive without any active human operation. Operate a motor vehicle. At present, automatic driving can be realized through sensor fusion. By using the image rear detection and depth detection including the entire rear of the vehicle, the position information of the obstacle in front in the three-dimensional space is estimated. [0003] However, current sensor fusion approaches require 2D rear end detection in images that include the entire rear end. So only when the ...

AI Technical Summary

Problems solved by technology

That is, when an obstacle partially enters the camera's field of view, the body (such as the front of the car entering the camera's field of view) or part of the rear of the car will not be used for image detection, which may result in only using the information of the depth sensor to estimate that part of the obstacle entering the camera's field of view Therefore, it is impossible to obtain the optimal estimated position of obstacles, resulting in defects in automatic driving performance.

Images