Design Of Automatic Leather Toggling Device Based On Vision Localization

With China’s rapid economic development,labor costs continue to rise,leather processing from manual production to automated production is the trend.Leather is soft and irregular in shape,and it is difficult to achieve automatic toggling by processing equipment,so the current leather dry finishing process of toggling drying process is still based on manual toggling.Manual toggling has low efficiency and worker labor intensity,which is difficult to ensure the processing quality of leather processing..To solve the above problems,this paper designs an automatic leather toggling device based on vision localization.The specific work consists of the following parts:(1)The overall design of the leather automatic toggling device.Firstly,the impact of the position of the gripping point and the spreading force on the change of leather area to determine the position of the gripping point and the size of the spreading force..The results showed that within the permitted range of tensile strength and elongation at break,8 spreading points are selected in the direction perpendicular to the backbone line,and 4 spreading points are selected in the direction parallel to the backbone line,a total of 12 spreading points,and the effect on increasing the area of the leather was more significant.Based on this result,this paper used the manual toggling process and action pattern to design the automatic toggling scheme.The basic mechanical structure of the leather automatic toggling device is initially determined,and an automated control scheme for the system is proposed.(2)The mechanical structure design of leather automatic toggling device.To achieve automatic leather stretching and quantitative stretching,this paper made use of the self-locking properties of the screw,gear meshing structure,linkage structure,and clutch structure to complete the design of the toggling clamp structure.This paper used the leather size and stretching experiments results to design the frame structure,and selected the screw and drive module stepper motor.According to the leather toggling drying process,this paper selected the chain transfer mechanism and hydraulic lifting support mechanism.According to the weight of the leather to be toggled,this paper selected the type of chain,Ⅳ three-phase stepping motor,and hydraulic cylinder.(3)Simulation and analysis of the key components of the leather automatic toggling device.This paper used finite element analysis software ANSYS WorkBench to simulate the toggling module and transport module key components for static and dynamic,the results showed that toggling clamp body,toggling clip block,linkage structure,gear mesh structure,clutch mechanism,and transmission module support feet,motor support plate were subject to the maximum stress in the material stress allowed range,that would not have an impact on the normal toggling device,indicating that design of the toggling module and transport was reasonable.This paper used kinematic simulation software ADAMS to analyse the movement of the toggling clamp,the results showed that the toggling clamp’s overall running speed is smooth,indicating that the design of the toggling clamp worked smoothly better.(4)Vision-based algorithm research on the positioning of leather edge gripping points.This paper used the conventional calibration method to simplify the calibration process and average it several times to calibrate the vision positioning system camera.By studying the existing image pre-processing methods,and comparing the commonly used image filtering methods,edge detection methods,threshold segmentation methods,and morphological processing methods,this paper used the median filtering method,maximum inter-class variance method,morphological open operation,and 8-directional chain code to extract the single-pixel contour of the image,and used the contour point-center-of-mass angle formula and normalization formula to filter out the target grasping points.The results showed that the average angle error of the algorithm is 0.16°,the average distance error is 0.56mm and the average running time is 32.48ms.Compared with the accuracy of manual toggling leather(manual gripping error is 1～3cm),the algorithm designed in this paper can achieve the localization of leather edge gripping points and meet the demand.