Algorithm Research On Embedded Hand Shape Recognition System Based On DSP

With the rapid development of Internet technology and information technology in twenty-first Century,the information security industry has also ushered in the industrial Peng hair period,the mechanization of traditional security technology has been difficult to meet the people's needs,as biometric recognition technology.Because of its individual differences and digital hiding,biometric identification technology has become a hot research direction in the new era.Hand shape recognition as a typical biometric identification technology,a collection of convenience,user acceptance,high stability,simple and easy to identify with other features such as the integration of advantages,has important research significance and practical value.From the point of view of industrial industrialization,the computer based hand shape recognition system has higher cost,larger volume,and does not meet the actual demand.In order to save the cost and enhance the practical performance,the embedded hand shape recognition system has a large market space,so this paper studies the hand recognition system based on DSP and the related algorithms.The main research contents are as follows:(1)The traditional hand shape recognition algorithm,by means of locating root and fingertip position of hand contour,with fingertips and refers to the root of its deformation caused the hand positioning is not accurate,hand shape feature extraction problem of error,the use of hand contour high stability,Voronoi diagram(Voronoi diagram)characteristics the whole algorithm can expression profile information,put forward a method of hand shape positioning algorithm based on Voronoi diagram.The experimental results show that the proposed method can effectively avoid the influence of the deformation area of finger tip and finger root on the accuracy offinger localization,thus improving the stability of the hand position.(2)This paper designs a kind of feature extraction and recognition algorithm based on the geometric characteristics of the finger,and the genetic algorithm of finger length interception coefficient and weight ratio to optimize the fingers,can effectively improve the recognition rate of hand shape feature.(3)This paper designs the hardware platform and software framework of hand shape recognition system based on DSP,developed the driver hand shape image acquisition based on DSP/BIOS and MATLAB platform in the design of location recognition algorithm good be simplified in accordance with DSP performance constraints,transplanted to the DSP system.The performance of the experimental platform is verified by experiments.