AGV (Automated Guided Vehicle) visual positioning system and method

Abstract

The invention relates to an AGV (Automated Guided Vehicle) visual positioning system and method. The system is formed by encoded points on the ground and a camera, a light source and an industrial personal computer which are on an AGV body, wherein each encoded point on the ground not only can provide unique encoded identification information and meanwhile, also provide information of global coordinates of a plurality of characteristic angular points; the global coordinates are acquired by a close range photogrammetry technology and are used for establishing a driving field electronic map so as to provide bases for AGV visual positioning; by calibrating a position and posture relationship between a camera coordinate system and an AGV coordinate system, accurate solution of a position and aposture of the AGV coordinate system under a global coordinate system can be implemented in the driving process on the basis of the characteristic information of the encoded points in a view field and a monocular visual positioning technology. The AGV visual positioning system and method provided by the invention have the characteristics of high flexibility, high accuracy, good real-time performance, high robustness and the like.

Description

technical field [0001] The invention relates to the field of robot positioning, in particular to an AGV visual positioning system and method. Background technique [0002] AGV (Automated Guided Vehicle, AGV) as an unmanned automated transportation equipment, it can carry a certain quality and run autonomously along the planned path, arrive at the designated place, and complete a series of predetermined tasks. It is used in the automated logistics transportation system, It has been widely used in the flexible production organization system. [0003] At present, the navigation methods adopted by AGV products at home and abroad mainly include electromagnetic navigation, laser navigation, inertial navigation, and visual navigation. Among them, the application proportion of electromagnetic navigation method is the highest, followed by laser navigation. The electromagnetic navigation method based on the principle of electromagnetic induction is the earliest successful applicatio...

AI Technical Summary

Problems solved by technology

For example, Qian Xiaoming et al. in the invention patent No. 201610286380.9 "Precise positioning method and system of monocular vision AGV based on multi-window real-time ranging" proposed to use the circular marker points in the far-end window, middle window and near-end window The geometric features of the AGV realize the precise positioning of the AGV. Due to the lack of coding information, the circular mark points used in this method cannot guarantee that the AGV can obtain continuous pose information during the movement process.

Hu Bin et al. proposed the use of two-dimensional code landmarks to realize the visual positioning of AGV in the invention patent "Method and system for locating the position of AGV working points by using two-dimensional code landmarks" with the patent number of 201710748607.1. The centroid point of the outermost square on the corner analyzes the position of the center point of the two-dimensional code relative to the center point of the image, and at the same time calculates the rotation angle of the two-dimensional code. Due to the lack of calibration of the camera coordinate system and the AGV coordinate system, only the deviation can be given. The navigation angle and the pixel deviation between the AGV and the two-dimensional code cannot give the global coordinates of the AGV; Liu Zheng proposed to obtain the spatial point through calibration in the invention patent "Indoor mobile robot positioning system and method based on two-dimensional code" with the patent number 201210186563.5 In the transformation relationship R between the image coordinate system and the car coordinate system, the deviation between the center point of the two-dimensional code and the center point of the image is extracted to realize the positioning of the robot. This method regards the camera as an ideal center projection model, which lacks camera optical center deviation and lens distortion. and other analysis, it is difficult to achieve high-precision robot positioning

[0005] To sum up, AGV visual positioning has not been widely used at this stage. The main difficulty is that the flexibility, real-time performance, robustness and measurement accuracy of the visual system cannot meet the needs of engineering practice.

Images