Research And Realization Of IC Card Ink-jetted Code Character Recognition Technology Based On Halcon

The modernized card production are all in assembly line-like large scale, In assembly line beginning has a mark-printing machine. In ink-jetted code on card, Many flaws may be product, such as partial characters are not printed, chaotic and position is not right, ect. So after ink-jetted code on card, they must be examined immediately in order to eliminate the wrong card timely to reduce or eliminate the losses card issuing companies. So it's necessary to research a kind of IC card character recognition system with high speed, precision and better robustness, which can make 100% reliable check on the character and symbol, achieving real "Zero /No Defects". This paper researches the Character Locating and recognition system in High-speed, including figure, letter and bar code, solving the problem such as low in contrast, noise interference, high range of ROI (Region of interest) variety, illegibility of the moving picture, low SNR(Signal Noise Ratio), the influence on productivity and the delay in picture treating, low precision and robustness, which can meet the online checking demand of package and card-making industry.Based on the above thought, we install the digital camera, which is connected w-ith the management microcomputer on the production line to gain characters printed on IC cards, also recognize them with some kinds of effective recognition algorithm. The thesis uses Halcon which applies universally accepted has the best potency machine vision software to design a new localization algorithm ,it carry on the character segmentation at first, then distill the goal character region, Use functions of Halcon to achieve improved neural network algorithms to identify characters and new Barcode identification and decoding algorithms.Through these algorithms, which can solve problems of character (numeral, letter and bar code) recognition and examination on high speed production line, satisfy technology requirements such as examination speed achieved to 30,000 cards per hour and recognitions rates of accuracy achieved to 99.99% , and break through the choke point in improving the speed of the manufacture process of thepackage and card-making industry, which improves the manufacture capacity of the equipment in this field and the competence in international market.The Paper integrate application of image acquisition, image processing, electronics, information transmission and storage of (the establishment of a database and visits), Halcon,ActivVisionTools,VC++.NET2003, knowledge and software development tools.