The Application Of Computer Techniques In Palm Print Image Processing

Dermatology is a subject which studies skin texture. Palm print includes a great deal of information. After extensive research, researchers have found palm print has something to do with nations, geology distribution, personality, temperament, health, intelligence and heredity. Especially the research results between palm print and disease(specially cancer) have shown a great future of palm print application. It will lay a good foundation for conquering caner. Much work of palm print research is done mentally at present. It will result in inaccurateness, slowness and uneasy popularization. There are no big and complete reference databases, consequently it will hinder palm print characters from being the indexes of medical disorders. how to deal with images intelligently and efficiently is very important to popularize the application of palm print research. It is necessary, therefore, to construct a robust, fast and flexible palm print information system. This paper states how to construct palm print information system in the following bearings: the difference between client/server and non client/server, what platform to use in server and client, system hardware platform, how to construct palm print database physically and logically, general design of application and so forth. The system takes advantage of computer hardware and software techniques and palm print study results. It makes most use of limited Image resources and converts analog images into digital images which can be used repeatedly and be transmitted via LANs, special communication networks and Internet. In Windows 2000, an Image Processing System is developed by the author using Visual C++ 6.0. Design idea, composition scheme, and modularized design method are also involved. Hilditch skeletonization can deal with any shapes, but it will take a long time to get the results. It needs repeating disposals. Close curve skeletonization needs a lot of 3 by 3 template and has 4 kinds of datas. Accordingly it is not easy to understand. Having compared the merits and unmerits of several palm print skeletonizations, a table-based lookup skeletonization is proposed. The methods is quick and easy to understand and program. Serveral palm print characters are proposed. Sixteen 3 by 3 templates are used to locate triraduises. More characters are calculated base on triradiuses. states extracting palm print triradiuses, the number of ridges between two triradiuses, is calculated using counter algorithm and outline line tracing. There is a lot of data in an image. It is not good for image storing and image transporting. In order to decrease space, speed transporation, an image must be compressed. There are two methods: revertible and unrevertible. the run length compression method is suitable for binary images.