Research On Identity Recognition System For Elderly Apartment

With the increasing problem of the aging population in China,the safety of the elderly has attracted people's attention.In order to provide a healthy,comfortable and safe living environment for the elderly,this subject focuses on the safety issues in apartments for the aged.Based on the facial recognition technology in the identification method,the identification system of the elderly apartment is designed and implemented under the platform of MATLAB.This paper studies the identification system of the apartment for the elderly,and designs a system based on face to judge the real identity of the old people.The identification system includes face image acquisition,image preprocessing,face detection,feature extraction and face recognition.First,this paper discusses the preprocessing method of face image.Because of the complexity of images in the elderly and the vulnerability of the face image to external influence in the process of acquisition,which increases the difficulty of image recognition.Accordingly,a series of adjustments on gray level transformation,equalization,edge detection,binarization and denoising,etc,are carried out to solve the problem of image recognition.Second,using the method of combination of skin color information and template matching to realize face detection.Skin color region image is extracted from skin color characteristics of the elderly and skin color clustering feature.Then,match it to the face template to get face region image.The simulation experiments on MATLAB platform shows that the proposed algorithm is effective and the detection rate is high.Finally,using PCA based identification method,and experiments under different conditions in ORL face database and self-built face database respectively.The experimental results show that the PCA algorithm has good robustness and stable recognition performance in ORL face database and self-built face database,and the recognition rate can meet the expected requirements of the elderly apartment.This paper contains the related technologies of face recognition in a series of identification methods,tested each part,and implemented the face recognition system based on MATLABR2014a platform.The recognition rate meets the requirements of the apartment for the aged.It has high stability and good application value.