Design Of Automatic Loading System Based On Visual Positioning And Measurement

With the development of domestic industrial automation,the automation system is gradually becoming an indispensable part of industrial production.In recent years,the production and processing of bags such as cement,flour and other bags has basically been automated,but their palletizes are still using a semi-automated operating mode,which requires artificial participation in the loading task of cargo.This operation mode not only has low efficiency,high work intensity,but also a harsh working environment is extremely harmful to the health of workers.Therefore,it is necessary to improve the automatization of loading system.Today,the development of machine vision technology has developed rapidly and is widely used in various industries,promoting the reform and progress of industrial production.In order to improve the degree of automation of the car loading system and solve the problems of inaccurate positioning of the carriage and the low efficiency of car loading,this article designed an automatic loading system based on visual positioning and measurement.According to the technical needs and functional requirements of actual industrial production,this article designed an automatic loading control system solution based on PC control terminals,which are inspected by the vehicle information collection module,carriage positioning and dimensional measurement module,and cargo position detection of the system scheme.The four modules of the coordinate calculation module and the loading palletizer execution module for function design and hardware selection.According to the system scheme,the software environment is established and functional to the automatic loading system.This article uses compartment positioning and size measurement design,cargo location detection and motion control of palletizer robots as the main line of research on the main line.Firstly,according to the needs of the carriage positioning and size measurement design,the camera is calibrated and the camera’s internal reference and distortion parameters are calculated.For the images collected by the industrial camera,there are incomplete vehicle images due to the shooting angle.This article uses the characteristic point extraction and matching of the SURF algorithm for the original image of the vehicle for the pre-processing of the vehicle.The FCN-ResNet50 network with a fusion CBAM attention mechanism is located on stitching vehicle images,and corner-point testing is performed on the carriage area.In the calculation of the corner-point image coordinates and spatial coordinates,this article proposes a triangle-based spacing algorithm based on dual-benchmark point,select the length of the pixels of the image unit of the base point,and calculate the image coordinates of the corner point of the carriage in combination with the benchmark point.The image coordinate and unit pixel length of the corner point calculate the space coordinate of the corner point of the carriage.The size of the carriage is achieved through the world coordinates of the corner point.Secondly,this article detects the location of the cargo through the algorithm detected by the YOLOX target,and calculates the position coordinates of the cargo grasp point with the triangular distance measurement algorithm.Combined with the size of the carriage and the size of the goods Analysis of target detection results.The D-H parameter model of the benchmarking robot is established with the six-degree of free palletizer robot to analyze the sports and working space of the palletizer robot.Finally,simulation experiment analysis is performed on the computing results of the compartment positioning and the calculation point of the cargo regulation,and the accuracy and stability of the automatic loading system is verified.