The Research And Implementation Of Business Card Recognition Based On Android System

With the rapid development of image processing technology, it has been widely used in more and morefields. The automatic recognition of business cards is just one of practical applications of imagetechnology. As a carrier of the individual information, the card is always in concise and beautiful andplays an important role in our business or normal activities. However, a large number of business cardsmay bring convenience to people while they cause trouble in daily management. Now a lot of Androidintelligent mobile phones and PDA can manage business cards, but few of them can realize theautomatic input or the effect is not ideal.Information extraction and identification is the subject of this paper. According to it, an automaticrecognition system for business cards based on Android operating system is developed. The course ofrecognition includes image preprocessing, page segmentation, character recognition, informationunderstands and so on. First, the current technologies and existing problems about business cardrecognition are described in detail and analyzed deeply. Then, a new binary algorithm particularly forbusiness card is proposed, which can remove interference of useless information to the maximumdegree and greatly improve the processing speed as well. After that, a combined algorithm based onconnected element and projection is used to analyze the layout. In dong so, the problems ofsegmentation and properties judgment of complicated layout can be better solved. As for the characterrecognition, a new fast recognition algorithm based on code matching is put forward which canimprove the recognition efficiency. Finally, the recognition and classification of information forbusiness cards are realized through heuristic classification rules of layout information.The research works in this paper includes almost all the key technologies of business card recognition.Experimental results show that the algorithms proposed in the paper not only can insure the recognitionaccuracy, but also can improve the recognition efficiency.