Research On Vision Guided Packaging Line Technology Of 4L Infusion Soft Bag

With the continuous improvement of living standards,people's requirements for their own lives and health are gradually increasing,and the development of the medical industry is also on the rise.Therefore,it should be paid close attention to the safety of medical supplies.At present,the packaging process of infusion soft bag is gradually moving towards mechanical automation.However,due to the softness and deformation of the infusion soft bag,it is difficult to fix them in the mechanical system,which leads to the difficulty of completing the mechanical automatic packaging process of infusion soft bag.In today's social environment and technological level,the packaging process of infusion soft bags is guided visually by the machine vision technology for detecting and judging the accuracy and stability of the packaging system,which has important application value.Based on machine vision technology,this paper designs a set of vision guidance algorithms for the packaging process of 4L large infusion soft bags.Based on the packaging steps,the accurate positioning,folding detection and heat sealing detection of infusion soft bag are mainly discussed and studied.The specific research work is as follows:(1)Firstly,the research significance and background of this subject are briefly explained,and the difficulties in the monitoring the infusion soft bag packaging are analyzed according to the actual situation.At the same time,the related theories and methods of image processing are studied for the tasks and problems needed to be completed,which includes image feature extraction,support vector machine and neural network,and actually uses these technologies to complete the vision guidance algorithm of the packaging process.(2)Secondly,according to the characteristics and packaging process of the infusion soft bag,an image acquisition system is designed to obtain the best quality image data,and the light source,lighting method and camera in the system are briefly introduced.In order to obtain the heat-sealing picture with stronger contrast effect,the lighting mode of using strip light source and lighting on the back of the seal was selected after experimental comparison.(3)Finally,based on the infusion soft bag sample images that are collected,the target area of the infusion soft bag in the image is extracted through operations such as image denoising and morphology,and features such as color and position are extracted and used to construct and train a support vector machine,which achieves the judgment for folding effect of the infusion soft bag.In addition,the heat-sealed area is extracted through Hough detection and learns to train a lightweight neural network for achieving the heat-seal detection of the infusion soft bag.Through the acquisition of image data of infusion soft bag and the multiple experiments and tests of the detection algorithm,the accuracy of folding detection algorithm is 99.9%,and that of heat sealing detection algorithm is 96.1%,all of which have achieved outstanding results.