Research On Intelligent Access Control System Based On Two Algorithms

Nowadays,with the accelerating development of science and technology,people's safety awareness is constantly enhancing.Within such context,an increasing number of biometric recognition technologies are applied to various industries,among which access control system is the most common security protection measure.Due to advantages like uniqueness and stability,biological features are widely used in identity recognition products.Therefore,comparing with traditional access control systems,the one supported by biometric identification technology is safer and more convenient.Among all kinds of biometric features,palmprint stands out in meeting the requirements of the access control system,since their features are stable,will not change under diverse conditions like aging,and are easy to be collected.The system will identify the user by palmprint if it is easy to collect;if not,for instance,when the user holds heavy stuff,the system then will utilize speech recognition technology and help the user pass through the door smoothly based on voice signals.Thus,this paper chooses two means of biometric recognition: palmprint and voice,to study the intelligent access control system.It is mainly divided into three aspects:(1)Palmprint recognition: palmprint recognition technology is briefly introduced.To recognize palmprint,there are a number of steps involved,including normalizing and sharpening processing for palmprint images,the application of Matlab for image binarization and denoising,and the preprocessing of the images by intercepting palmprint' corresponding regions of interest(ROI)according to their contour information and finger root locations in the contour chain code.Finally,different types of algorithms are discussed while 2D-Gabor is selected to extract palmprint features to realize the palmprint recognition.(2)Speech recognition: the speech recognition technology is summarized.To recognize speeches,the system needs to preprocess voice signals through framing and windowing,and adopts methods of endpoint detection,feature vector extraction,model training,dynamic time warping and similarity measurement.This part also introduces the training method of speech templates,and the way to optimize the DTW algorithm.(3)Hardware and software designs of the access control system: regarding palmprint recognition,image collecting device is selected,and by using images from the image acquisition toolbox of Matlab and functions of GUI,an on-line access control simulation system based on palmprint verification is designed.In regards of speech recognition,peripheral hardware circuit is constructed,SPCE061 A SCM is selected as the main control part,and the programming is completed.Both recognition methods are tested and the results are analyzed.Experiments show that the access control system has a high recognition accuracy and basically meets the expected requirements of the system.The idea of combining two recognition methods has great value in terms of the design and research of intelligent access control system,and has widespread practical use and market prospects.