Alignment system and method for automatic charging of robots based on vision

Abstract

The invention discloses an alignment system and method for automatic charging of robots based on vision. An active visual guidance device is installed on a charging base, a common camera is installedat the charging end of a mobile robot, the position opposite to the charging base and an approximate angle of the robot are calculated according to the position of a guidance marker in an image captured by the camera and visual characteristics, and then the robot is accurately aligned to the charging base by adopting an alignment motion control algorithm. Positioning markers such as magnetic barslaid on the ground or two-dimensional codes are not needed, so that the problem does not exist that failure is caused by manual or mechanical abrasion. The calculation of the relative position of themobile robot can be up to the precision of image pixel level, real-timeliness is good, and the precision is high. Compared with an image identification scheme of a charging base body, the active visual guidance device actively transmits marking signals, is simple in identification, has higher signal to noise ratio and can be obviously different from a background, and therefore guidance and positioning are more stable and reliable.

Description

technical field [0001] The invention relates to the technical fields of computer vision and robot automatic charging, in particular to a vision-based robot automatic charging alignment system and method. Background technique [0002] With the increasing application of mobile robots (such as AGV), solving the problem of automatic charging of mobile robots is the basis for ensuring their long-term work, and it is also a must-have function for mobile robots. The core technology to realize the automatic charging function is to realize the automatic alignment of the charging electrode of the robot and the charging stand of the charging pile. seat, and then perform docking. The alignment process requires high precision, the allowable error is within the centimeter level, and at the same time requires a relatively fast speed to improve the working efficiency of the robot and avoid the extreme situation of battery exhaustion during the alignment process. Therefore, the alignment t...

AI Technical Summary

Problems solved by technology

Infrared guidance is low in cost and simple to implement, but its positioning accuracy is not high

[0006] 3) Laser guidance method (such as Patent No. CN201310567061.1 "A dual-guided robot autonomous charging method"), paper "Indoor Navigation Algorithm and Autonomous Charging System Research for Hospital Service Robots", Luo Zhenhua, Harbin Institute of Technology, paper "Indoor Environment Research on Autonomous Charging of Mobile Robots", Hao Zongbo et al., Journal of Harbin Institute of Technology, 2005) use laser radar as a sensor to detect laser signals, and determine the position and angle of the robot relative to the charging base by finding and matching the contour features of the charging base or other marks. The hardware cost of using laser sensors is high, which is not conducive to the cost control of mobile robots and the promotion to the consumer market

[0008] a. Solutions based on QR codes, such as patent number CN201720386953.5 "A highly reliable and power-saving automatic guidance system for AGV robots", the paper "Research on Flexible Docking Charging Technology for AGV Cars", Zhou Wenjun, Light Industry Science and Technology , 2015, using the two-dimensional code visual guidance method, it is necessary to add a two-dimensional code identifier on the ground as a marking point for robot positioning. The disadvantage is that the two-dimensional code is easy to blur or wear after long-term use, resulting in difficulty in recognition

[0009] b. For the image recognition scheme of the charging stand body, such as the paper "Inspection Robot Autonomous Charging and Docking Control Method", Wu Gongping, Journal of Harbin Institute of Technology, 2016, this kind of method recognizes the charging stand in the image through visual methods, and then calculates The position of the robot relative to the charging stand is realized to achieve alignment control. This kind of solution is relatively sensitive to changes in light and environment, and it is not convenient for different types of charging stands.

Images