Double-vision tracking device and method used for omni-directional movement robot

Abstract

The invention discloses a double-vision tracking device and method used for an omni-directional movement robot. The tracking device comprises a front end vision tracking system installed at the front end of the robot, a rear end vision system installed at the rear end, and a movement controller. A vision recognition system provides the guide and judgment foundation for movement of the robot by acquiring front route images and rear route images of the robot and extracting offset distance and deflection angle information. The front end vision tracking system and the rear end vision tracking system are each in communication with the movement controller and transmit information of the deflection angle and the offset distance in real time. The movement controller calculates the optimal movement route information of the robot and then regulates the speed of each wheel of the robot so that the front end and the rear end of the robot can both move along the route. The deflection between the front end of the robot and the reference route and the deflection between the rear end of the robot and the reference route are each detected through the two-way vision recognition system, more complete route information is obtained, the control precision is improved, the tracking posture of the omni-directional movement robot is determined, and better insurance is provided for movement navigation and obstacle avoidance of the robot.

Description

technical field [0001] The invention belongs to the field of robot tracking and path recognition, and in particular relates to a dual vision tracking device and method for an omnidirectional mobile robot. Background technique [0002] When the path is known, fast and accurate path tracking is a very important problem in mobile robot applications. In the process of path tracking, the path is often recognized by robot vision and path information is extracted. However, ordinary one-way vision can only monitor the path deviation of the front end of the robot, but cannot detect the rear end, which may easily lead to a decrease in tracking accuracy or even a collision accident. Two-way vision can detect the deviation between the front and rear ends of the robot and the reference path, and can track the path with high precision and make the movement safer and more stable. [0003] Traditionally, single vision sensors are mostly used. Compared with ordinary one-way vision, two-way...

AI Technical Summary

Problems solved by technology

The phenomenon of "tail flicking" is the situation where the front end of the robot is on the guide line and the rear end deviates far away, which will cause the tracking accuracy of the robot to decrease, and may even collide with obstacles or walls

Images