Research And Design Of Online Identification Authentication With Palmprint Based On SOPC

Biological recognition technology makes the use of inherent biological characteristic of human body to recognize identification automatically. Biological characteristic of human body is the features of universality, unique and stability. Compared with the traditional identification recognition method, the biological recognition technology can caters to the demand of security and convenience of human identification authentication in the modern information social. There are a lot of human biological characteristics to recognize identification, such as fingerprint, iris and person face. Biological characteristic recognition technology with Palmprint is a new research way in recently years. Palmprint recognition which was concerned as an important complementarity of biological recognition technology has some advantages: palmprint contains rich information; the method of gathering the palmprints is simple, so the cost of equipment is low; the main character of the palmprint is stabile and distinct ,not interfered by yawp, so the precision of recognition is high.The application domain of palmprint recognition is wide. Except the requirements of system algorithm reliability, Palmprint recognition also demand equipment smalle, convenient and portable. Therefore, escaping the fetter of PC, embedded system of palmprint recognition has more practical application value and wider application domain. But now, the technology of embedded palmprint recognition system has not far from developed sufficiencily.Referencing on domestic and overseas relevant study, and in the experiment condition of small scale sample, the paper extract the main line characteristic of the low resolution image of palmprint, which is self-contained to the palmprint recognition, and design embedded palmprint identification authentication system based on the technology of SOPC(System On Programmable Chip).With the algorithm design, the paper extract the main line characteristic of palmprint by combining image process and mathematical morphology. And it defines characteristic space by polar coordinate modality. At last we sue Fisher criterion to fall characteristic space dimension, completing the last part of confirmation algorithm by neural network.Applying the technology of SOPC which is based on the embedded processor-Nios, the system structure with the core of NiosⅡwas designed. The hardware design including Nios embedded process, image capture module, Matrix keyboard, LCD display electric circuit and GSM communication module of Siemens is designed by the help of development tools: QuartusII, SOPC Builder, IDE and the program language of C and C++. At the same time, we finished the software interface design of all modules, and applied REF network model to realize on-line identification recognition.The whole system integrated confirmation algorithm, system model and interface function into a single FPGA chip. The system has obvious advantages: algorithm renewing ,software and hardware are reconfigurable, high reliability, small volume etc. Testing result shows that the system can meet the demand of security rank basically in populace domain, and the requirement of real-time and reliability of the on-line identification recognition.