Research On Key Technology Of Automatic Identification Of Billet End Face Information Code

Character recognition technology is an important field of pattern recognition application.Through the use of computer technology,a large of words or numbers in people's production and life were processed,and intelligent recognition results were obtained.With the comprehensive development of China's informatization construction,artificial intelligence has entered various fields,and intelligence has been increasingly valued and respected by people,such as the identification of license plate numbers in the transportation industry,fingerprint recognition at work,face recognition,document identification in life,ticket identification and barcode recognition,etc.Due to the unique intelligence of this technology,there are huge market demands for industrial applications.Therefore,this paper mainly studies a character recognition technology that is applied to the artificial steel in the steel field.This paper mainly studies three aspects: the location of the information code,the inclination correction of the continuous casting billet's face information code and identification information recognition algorithm.The main work and innovative research include:1)Based on the recognition principle of the end face information code of continuous casting billet,the automatic recognition system was designed,and the working principle,selection of hardware equipment and system construction of the system were studied;2)According to the workshop environment and the inherent characteristics of the continuous casting blank end face,the effective character information in the collected images was difficult to be separated from the background environment,Aiming at this problem,a method combining corner detection and expansion method is proposed,which can effectively extract the coordinate position of the closed area of square steel,the valid character area was accurately located;3)For the actual situation on the billet conveying roller,the billet end face has a vertical and inverted inclination.The correction method combining the gravity center difference feature and the horizontal analysis method is proposed to solve the problem of tilt correction;4)In view of the character inclination of the round bale end faces,a method for extracting the coordinates of the image surface of the round bale end faces was developed.The image plane coordinates before and after the tilt were calculated,and the initial Angle correction was completed.A correction algorithm for counting the number of vertical projection columns and the minimum value was proposed,and the final correction of the arbitrary inclination angle of the round end face character was completed;5)For the coding of the handwritten information,the method of step-by-step segmentation was proposed,and the problem of character sticking was eliminated;6)In the process of information code recognition,an algorithm is proposed to extract image feature information and describe the feature parameters to match,to ensure the accuracy of identification;7)In order to realize automatic identification,The design and development of the software of automatic recognition system for the end face information code of billet with MATLAB was completed.Finally,the feasibility of the content of this paper was demonstrated through experiments.In this paper,the system and method of billet end information code identification are presented,which can accurately identify the steel billet end information,and lay a foundation for the information identification in the manufacturing process.Experiments show that the end face information code identification method can meet the tracking information of billet.It provides the early theoretical basis and technical support for the automation of the tracing operation of billet end information,the improvement of accuracy and production line efficiency.