Facial image based security system using PCA

Security is one of the utmost requirements of homes and businesses which requires biometric identification. This paper aims to identify a person through face recognition. Face recognition is very complex and multidimensional problem. A MATLAB based Principal Component Analysis is used for face matching decision. The system is designed in MATLAB which converts facial images to feature characteristics of initial training database images. Facial features are extracted from the face and eigenvalues are calculated and represented as a eigen vector. Comparison of unknown face image and database image is done using Euclidian distance method. The recognized facial image has minimum Euclidian distance with the database images. When face is recognized by MATLAB Code it will provide the signal to ARM7 to open Gate and System will send SMS to the authorized person using GSM module. Security system using MATLAB and Embedded system design is cost effective, reliable and highly accurate.