Embedded Low-power Partial Fingerprint Recognition System Design And Implementation

Embedded fingerprint identification system is the digital collection of the fingerprint,the means of digital image was used to extract fingerprint information,through certain recognition algorithm to complete matching,finally use fingerprint to complete identification function.Because the fingerprint algorithm in digital image processing requires a lot of multiplication and calculation,so normally driving high-performance recognition system on the hardware or the high performance DSP chip ARM series processor chips,but inevitable limits the overall energy consumption on certain level,the system of the life of the battery power supply system is bound to have much impact.The paper puts forward a kind of in the dual structure of high-performance low-power embedded fingerprint identification system is designed,in order to we can to achieve including registration,delete and certification,etc,with the ARM kernel and ML67Q5250 processing chip with the circuit of hardware accelerators to use as the main control module of identifying fingerprint,and combined with the adaptive optimization algorithm of fingerprint.Combined with a at the same time we have ultra low power 16-bit microcontroller chip MSP430f159 can be used to control button,display,motor drive,etc.,and other peripheral modules,finally satisfied we make sure the system has low power consumption,and can control the fingerprint recognition module in fingerprint recognition module time start the electric energy,reduce energy consumption of the module work time,thereby reducing energy consumption.At present,many research institutions have extensive and in-depth research of core technology and science and technology in the field of fingerprint identification,synthesis of algorithm on good when fingerprint image processing effect is better,both in terms of speed or accuracy can be very good guarantee,but for the low quality of incomplete fingerprints,because of the dirt of fingerprints,scar,fracture,etc.,make the image appear lost and nonlinear serious deformation,so a lot of difficulties in the identification.According to the fingerprint of archives information,the proportion of the excellent fingerprint is limited,the fingerprint of incomplete proportion is not allow to ignore.Above all,how about the broken fingerprint accurate,efficient,a fingerprint direction categories is an urgent need to solve the difficult problem.The author's main work is as follows:(1)feature extraction,by looking for fingerprint matching degree and and in the end,through the four aspects of fingerprint indexing to seek a new way of optimization and algorithm,and in a deeper level,research how to promote incomplete fingerprint image enhancement methods.(2)the discussion is based on information entropy is the reconstruction of the fingerprint of the broken part and correction algorithms,this system is designed in according to the characteristics and the details point information integrated and combined with the fingerprint direction field estimation,on the basis of comprehensive in accordance with the need of image enhancement of partial fingerprint,the introduction of more accurate estimate,and is ultimately based on the direction of the field of field information for the direction of the broken fingerprint region to a more comprehensive calculation.(3)set up the control of peripherals module for the center with MSP430F149,contains data storage module,calendar function module,keyboard input module,motor driver module,etc.,as a result,we can use this function as an electronic password,and efficacy of the system to achieve the effect of the minimum.