The Study Of Fingerprint Identification System Based On ARM7

With the development of the society, science and technology are more and more advanced,and the level of the people's lives is improved.But at the same time,the sec- urity issues are increasingly serious. The previous identification technology cann't me -et the demand of the society,for example the password,the inked fingerprint and so on.For these reasons,the identification technology based on biology comes forth.The Biometric Identification Technology refers to biometrics as the representa- tive of a person. So the safety factor is higher. As there are many advantages of the fingerprint such as unique, more portable , not easy to be copied, and invariance all through the life, the Fingerprint Identification Technology is the most skillful, the pop -ular and practical one compared with the other Biometric Identification Technology. Moreover, with the development of technology, the components are smaller and smaller with higher-performance, lower power, lower price,and the fingerprint image processing algorithms are increasingly sophisticated and accurate.These allows the fingerprint identification system can be widely applied.At present, fingerprint recog- nition technology has been used in the attendance checking, banking, security check- ing and other fields.In order to improve the performance of fingerprint, the system uses a sweeping fingerprint sensor AES2510.AES2510 is one of the capacitive fingerprint sensor, which uses the RF Technology.It collects true fingerprint and is hanrdly affected by the status of the finger. It won't make errors because of the dirty fingers,the shed fingers or the wet fingers,Etc.The fingerprint latent image left by the user can no more be utilized by the criminal. The system uses S3C44B0X which manufactured by Samsung as the embedded processor . S3C44B0X is of ARM7 series ,and it has a lot of advantages such as small,powerfull,low consumption and low price. The finger -print identification system based on S3C44B0X processor and fingerprint sensor AES2510 can perform automatic fingerprint identification,and the Performance price ratio is higher.These make this system can be popular used in people's daily life, such as household door locks, attendance machine, ATM machine,and it has a broad market potential.The article first discusses the principle of fingerprint recognition technology and its advantages compared with other Biometric Identification Technology. It outlines the level and trends of the development of fingerprint recognition technology at home and abroad, and predicts its future prospects. It also describes the overall design and the process of the system: After the power-on reset, by initializing the fingerprint sensor, the host makes it enter fingerprint detection state.When detecting a finger,the sensor will enter imaging mode and collect the fingerprint,then the host read the data from the sensor.If it is to register the fingerprint information ,the data can be stored in the SD card, providing sample data for the match,and at the same time, with two small LED lights showing whether the registration is succeed. If it is for identification, it should make the fingerprint comparison, and then according to the results control the operation of the lock or alarm.Second, the article studys each part of the minimum system,including power supply design, the clock circuit, reset circuit, JTAG debug circuit, SDRAM and FLASH memory expansion circuit. Later it discusses the interface, the internal registers and the commands of SD card, and also describes the processes when writing,reading or erasing the SD card . The fourth chapter details the hardware and software design of the fingerprint collecting module, including the working principle of the fingerprint sensor, the sensor and controller interface design, the software design of the fingerprint collecting,registration,matching and the image processing technology. The purpose of image stitching is to synthesize the image fragments transmitted from the sensor into a complete fingerprint image. Then, it describes the fingerprint image processing algorithms during the fingerprint identification process. This is because The fingerprint image must be preprocessed before registed or matched.The image preprocession including image normalization, the effective region extraction, image enhancement, binarization and thinning,is to facilitate feature extraction. Image normalization making the mean and variance of gray values close to the given value, is to make the following process easy . To compress the data, and improve the speed of the system, and we should extract the effective area of the fingerprint. Image enhancement removing noise from the image is to improve image quality and to improve the correct-recognition rate. Binarization transforms the gray level image into a image only having black and white values, image thinning reduces the pixels of the ridge edge to make it only one-pixel- wide. Both of the pprocessing can reduce the amount of data ,and are to improve the speed of the system, but also to facilitate follow-up feature extraction. After the image preprocessing,it must proceed the feature extraction, and then the system can implement fingerprint identification. Finally,it describes the purpose and method of the system test.Then analyzes the experimental results and comes to a conclusion.This paper describes the main part of the system design,and the experiment has completed the hardware design and welding .Besides, it has debugged the main modules of the system.These make up of the firm basis of the further study, and someome who designs a fingerprint identification system based on S3C44B0X or fingerprint sensor AES2510 can also refer to this paper.