Research On The Detection And Location Algorithm Of Storage Box Based On Contour Feature

With the rapid development of modern logistics technology and the rise of labor costs,it is more and more difficult for traditional warehousing logistics technology to meet the needs of logistics and distribution.Artificial intelligence-supported warehousing logistics robot technology has begun to be applied to the field of warehousing logistics.The robot’s efficient and real-time box cargo detection in the warehouse has become an urgent problem to be solved.This paper intends to extract the contour characteristics of the box body through visual sensors,realize fast,stable and accurate box goods identification and positioning in the warehouse scenario,solve a series of operation automation problems such as picking,palletizing and handling of box goods,and complete the relevant Software algorithms and software systems.The main research contents of this article are as follows:(1)According to the needs and problems of the storage box detection,the corresponding error index is formulated,and the system scheme of the storage box identification and positioning in this paper is given.Firstly,the requirements of the loading and unloading mobile robots in the warehouse environment are analyzed,and then the sensors are selected for the problems in the box cargo identification research at this stage,and the overall system plan of this paper is given.Identify the positioning plan for segmentation and description.(2)The improved straight line feature detection algorithm is studied,and a box cargo identification method based on contour features is proposed,which improves the robustness of the box identification algorithm.Considering that the rectangular outline of the box is the most stable feature,it is proposed to realize the identification of the box goods by detecting the rectangular outline.Aiming at the problems such as the lack of straight line detection and the edge corner break in the original straight line detection algorithm,an improved straight line detection algorithm based on iterative verification is proposed.Based on this,this paper proposes a box cargo identification method based on improved straight line detection algorithm,and verifies its feasibility through experiments.(3)The experiment analyzes the ranging model of the depth camera,and proposes a calculation method to solve the box posture and size based on the contraction feature points.In terms of box cargo positioning and three-dimensional measurement,the ranging principle of the depth camera is introduced,the depth measurement accuracy is analyzed experimentally,and the internal parameter matrix and distortion coefficient of the depth camera are calibrated.Finally,a specific calculation method for solving the position and size of the box cargo through feature points and feature planes is given.(4)The mobile robot platform has been developed,and experiments have verified the feasibility of the identification and positioning algorithm of the warehouse box goods in this paper.Designed and built a mobile robot platform,transplanted the designed box cargo identification and positioning algorithm into the robot system,completed the robust experiment of the box identification algorithm and the error experiment of the position and pose solution algorithm,and this paper verified the feasibility of the proposed case goods identification and positioning method.Experiments show that the algorithm in this paper has good real-time and robustness.In summary,in view of the needs and problems of storage box detection,this paper proposes a method for identifying and positioning storage boxes based on contour features.The line contour is extracted by an improved straight line detection algorithm,and then the box is calculated by shrinking feature points.The posture can realize the fast,stable and accurate identification and positioning of the storage box,which is suitable for the detection of the box in the storage environment such as the intelligent warehouse.