Research On Intelligent IC Card Defect Detection Based On Character Recognition

With the large-scale production of modern IC cards,in the process of imprinting characters at the front end of the assembly line,it is inevitable that some characters such as missed printing,missing,and confusion will appear.Therefore,it is necessary to carry out real-time identification and detection of the embossed characters.In order to reduce the loss of the issuing company,the company will eliminate the defective card.Traditional detection methods such as manual detection and template matching are low in efficiency and poor in practicality and cannot meet the needs of intelligentization.In order to identify the high reliability of IC card imprinted characters,this paper studies the technology of IC card imprinted character recognition and detection.It can detect the characters' reliability and achieve the IC card character recognition system with fast detection,high accuracy and good robustness.This subject first analyzes the characteristics of the IC card visual imaging,the type of noise in the image and the solution to the denoising.It divides the IC card character image and increases the edge features of the image,which lays the foundation for the subsequent recognition of the positioning character.According to the imaging characteristics of the embossing characters,the location and recognition of the embossing characters in high speed motion are mainly studied,and the technical problems of the character edge detection,the low signal to noise ratio,the poor contrast of the image,the selection of the ROI region and the image segmentation are solved in the high-speed movement.At the same time,the edge breaking often appears in the character contour extraction.The two step graph is used as a template,and the boundary tracking is used as a tool to repair the edge of the fracture.The result shows that the method is suitable for the broken edge character.Based on the above ideas,by selecting appropriate camera and lens and other hardware facilities to obtain the IC card character images,and build a test platform,the software part of the entire platform is written in M language of MATLAB.After the character recognition test of 1000 card samples,the results show that the IC card character recognition system can effectively realize character recognition,and the overall recognition rate is 99.8%,which meets the requirements of IC card character recognition system.