Obstacle recognition method and device, terminal, and computer-readable storage medium

Abstract

The embodiments of the invention provide an obstacle recognition method and device, a terminal and a computer readable storage medium, which relate to the field of local navigation of robots and intelligent vehicles. The method comprises the following steps: obtaining a 2D coordinate set of the surrounding environment through radar, and obtaining a 3D coordinate set of the surrounding environmentthrough a somatosensory camera; adjusting the 2D coordinate set and the 3D coordinate set according to a preset algorithm to obtain a target coordinate set, so as to unify the abscissa values and theordinate values of the 2D coordinate set and the 3D coordinate set; and recognizing an obstacle according to the target coordinate set. The real-time performance and accuracy of obstacle recognition can be maintained, and the overall characteristic of an obstacle can be displayed.

Description

technical field [0001] The invention relates to the field of local navigation of robots and intelligent vehicles, in particular to an obstacle identification method and device, a terminal and a computer-readable storage medium. Background technique [0002] In the automatic driving of mobile robots and various vehicles, the detection and recognition of obstacles occupies a very important part. In order to better realize automatic driving, the detection and recognition of obstacles needs to maintain real-time and accuracy, and at the same time, it needs to express the overall characteristics of obstacles. At present, the obstacle detection and recognition of existing unmanned equipment usually adopts a single radar or a single vision method. [0003] However, when a single radar is used for obstacle detection, if the radar used is a 2D radar, since the data collected by the 2D radar is at the same height and plane, it is easy to make a wrong judgment on the type of obstacle....

AI Technical Summary

Problems solved by technology

[0003] However, when using a single radar method for obstacle detection, if the radar used is 2D radar, since the data collected by the 2D radar is at the same height and plane, it is easy to make a wrong judgment on the type of obstacle

If the radar used is 3D radar, it can solve the problem that 2D radar can only collect at the same height and plane, but the price is too expensive

When single vision is used for obstacle detection, the real-time performance of obstacle recognition cannot be maintained due to the lag of the camera

Images