Tray posture identifying system for AGV forklift and method thereof

Abstract

The invention discloses a tray posture identifying system for an AGV forklift and a method thereof. The tray posture identifying system comprises an image acquisition module for acquiring the originalimage of a tray, an image identifying module for receiving the original image obtained by the image acquisition module and identifying a key position area in the original image, a data acquisition module for obtaining the tray central position in the key position area, the left side height h1 of the tray and the right side height hr of the tray in the key position area, and a data calculating module for calculating the offset and the rotary amount. The tray posture identifying system has the following benefits: the AGV forklift in the tray posture identifying system and the method thereof canautomatically identify the key position area of the tray in the image and can automatically identify the rotating and the deviating degrees of the tray relative to the self from the key position areato obtain the posture information of the tray relative to the AGV forklift.

Description

technical field [0001] The invention relates to a pallet posture recognition system and method for an AGV forklift. Background technique [0002] With the rapid development of modern e-commerce and logistics technology, AGV (Automatic Guidance Vehicle) is playing an increasingly important role in intelligent storage technology, and it is crucial to the detection and identification of cargo pallets. [0003] When the AGV forklift wants to fork the pallet, if the angle between the pallet and the AGV forklift is greater than a certain threshold, or there is a large offset between the position of the pallet and the forklift, the fork of the forklift cannot be inserted into the bottom of the pallet smoothly. . Therefore, it is necessary to recognize the pose of the pallet in order to adjust the pose of the AGV, and then realize the smooth forking of the pallet. [0004] The current detection of cargo pallets mainly uses visual detection and lidar detection. Although the cost of...

AI Technical Summary

Problems solved by technology

[0004] The current detection of cargo pallets mainly uses visual detection and lidar detection. Although the cost of monocular vision is not high, the accuracy is not high, and it is greatly affected by light; multi-line laser and binocular vision detection have the characteristics of high precision. However, multi-line lidar is expensive, and double-sided visual detection is greatly affected by light; single-line lidar is low in cost, but multi-line lidar is too expensive, and double-sided visual detection is greatly affected by light.

Images