Study And Application Of Bar Code Identification Method Based Pattern Recognition Technology

Current situation of oilfield material management,dependents on material management system,and the most important part of the material management system is the barcode scanning function.Due to the weather and the environment in the field,it result out the low efficiency of material management.Especially,if the tools are so many,it often affects the flow of tools severely,and lead to the loss of enterprise.So,this paper improves barcode identification technology based on pattern recognition technology,analyzes and solves the problem that the bar code image can not be identified because of tilt,noise and dirt.By simulation and practical test,the design can improve the bar code recognition technology effectively,and make the management of oilfield materials more automatic and efficient.The main research in this paper is as follows:(1)Hough transform image tilt correction:for the oblique bar code image by scanned,this paper can solve the problem that the barcode can not be identified because of the skew of the image,based on Hough transform image tilt correction method.(2)Improving median filter algorithm based on support vector machine:firstly,for the noise problem of bar code image,the median filtering algorithm is used to treat barcode images,and the image noise can be removed.It can remove the black and white isolated noise points on the bar code image,by the median filtering algorithm.Then,for the advantages and disadvantages of the median filter algorithm based on pretz-salt noise characteristics,the advantages and disadvantages of image filtering algorithm for support vector machine and the characteristics of pretz-salt noise,the standard median filter algorithm are improved,this paper improves the median filter algorithm based on support vector machine.Simulation experiments show that the algorithm can eliminate pretz-salt noise more effectively and have higher SNR.(3)Ostu algorithm based on adaptive threshold selection:firstly,for the smudge problem of images,this paper adopts Ostu algorithm and Bernsen algorithm to process images.According to the actual situation of oil field,comparing the processing effect and running time of these two algorithms,this paper uses the Ostu algorithm to handle bar code images.Then,for the operation time of Ostu algorithm can not meet the needsof field material management,the Ostu algorithm is improved.In this paper,the adaptive threshold selection method is used to improve the Ostu algorithm.The improved effect analysis shows that the improved algorithm can speed up processing time and improve productivity.