| Machine vision is an interdiscipline involving many other fields.Compared with traditional artificial inspection,visual inspection has the advantages of high detection accuracy,high efficiency,safety,low cost and non-destructive inspection,therefore,machine vision has received great attention in the industrial field.This paper firstly analyzes the background of the research topic,and we have a full understanding of the research status of machine vision technology in the field of pharmaceutical packaging after reading a large number of literature,and aiming at the shortcomings of the current visual inspection technology in the field of drug packaging,the algorithms to improve the detection accuracy and efficiency were studied,and a visual inspection system platform was built to simulate the industrial production line.A complete set of information machine vision software system for phases III of medicine has been developed based on C# and HALCON,The system has been used in mass production.The research work and innovation of this paper are as follows:(1)In order to improve the detection accuracy of the visual inspection system for medicine information,a delay-based misplaced target image difference denoising algorithm is put forward in this paper.In order to eliminate the impact of printing and packaging noise on detection results,as images acquired are characteristic of delay,the infinitesimal displacement between characters acquired at two times can be used to process three-date characters after the difference is solved in the image enhancement algorithm to promote the three-date image denoising effects in drug packaging.(2)A delay-based misplaced difference subtraction denoising algorithm is designed in this paper.Drug packaging images can be denoised by superposition of two images with delay.To be specific,as noise is a random occurrence in the background image,two images with delay are added together,a threshold is designed to eliminate the background noise,and the coverage of three-date characters in space is increased to enhance the accuracy of character recognition.(3)In order to improve the detection efficiency of the visual system,a high speed medicine packaging detection technology based on real time image identification is proposed in this paper.Firstly,the medicine packaging image with "three phases" information is preprocessed to generate the optimal image dimension reduction matrix operator.In the process of drug packaging detection,the image data collected by the host computer system is directly stored in the system memory to form a queue of images to be processed.The system follows the pre-trained image reduction dimension operator matrix,and takes the chain group processing method into the memory.The image data is subjected to dimensionality reduction processing,and character recognition and product integrity determination are performed on the dimensionally reduced image.The visual inspection system packages the qualified medicine according to the result of the medicine packaging identification,deletes the original image to reduce the memory occupation,and starts the storage thread for the package image with the problem,and saves the image for post-mortem inspection and analysis;(4)A complete three-phase information visual inspection system is built based on C# and HALCON in this paper,which combines theory with practice,uses C# source code to implement the improved algorithm of this paper,and embeds it into the visual system.In order to improve the operating efficiency of the system,The system framework and algorithm part are developed separately,and the detection system has been mass-produced and developed.The machine vision detection system for medicine information developed in this paper has a detection accuracy of 97.8% through experimental tests.Based on the improved detection system,1136 images can be processed per minute,and the detection accuracy and efficiency of the visual detection system have been significantly improved. |
PDF Full Text Request