AGV pedestrian following method based on monocular camera

Abstract

The invention relates to an AGV pedestrian following method based on a monocular camera. The method comprises the following steps: 1) detecting a pedestrian target: obtaining a pedestrian target detection frame according to a pedestrian detection model deployed on an upper computer; 2) using a monocular camera to calibrate a homography matrix: obtaining a homography matrix H from a three-dimensional world coordinate system to a two-dimensional pixel coordinate system; 3) pedestrian coordinate calculation: calculating coordinates of a contact point of the pedestrian and the ground in the three-dimensional world coordinate system, namely world coordinates (xw, yw) of a pedestrian target; 4) calculating the speed and the angle of the mobile robot in real time: respectively designing a linear speed PID (Proportion Integration Differentiation) controller and an angle PID controller to calculate the linear speed and the angular speed of the mobile robot in real time; and (5) controlling the chassis to move, specifically, enabling the upper computer to issue linear speed and angular speed information to the lower computer through an ROS system, enabling the lower computer to resolve a speed instruction into the expected rotating speed of a driving motor, so that the AGV follows pedestrians. Compared with the prior art, the invention has the advantages of simplicity, high efficiency, real-time accuracy, distributed communication of the master and slave machines, small calculation amount and the like.

Description

technical field [0001] The invention relates to the field of robot target detection and tracking control, in particular to an AGV pedestrian following method based on a monocular camera in an indoor environment. Background technique [0002] The rapid development of computer vision-related technologies and the improvement of hardware computing speed have greatly enriched the functions of the robot. The mobile AGV service robot uses a monocular camera based on ROS to easily and efficiently estimate the distance and angle of the target pedestrian, and maintain a certain distance. The safe distance will have a wide range of application value in office, inspection, welcome and other scenarios in the future. [0003] The existing technologies for distance estimation of pedestrian targets using sensors mainly include the following: [0004] One is the ranging method based on laser radar: the common method of laser ranging is triangular reflection method. Triangular ranging is ba...

AI Technical Summary

Problems solved by technology

However, the defect of this method is that the camera baseline limits the measurement range, and the accuracy of time measurement is high, and the accuracy of the estimated distance of the depth camera is difficult to reach the millimeter level, and it is impossible to accurately measure black objects, transparent objects, and close-range objects.

Images