Abstract: Since lot of forgeries occur in our day to day life, security and surveillance has become a mandatory aspect. Over the last few decades, face recognition has become a popular area of research in computer vision. This system may be implemented in real time systems requiring user authentication such as attendance systems, ATM security, Electronic Passports and Network security. Our project aims in developing precise way of detecting faces using an algorithm called ”Eigen face detection” which uses Principal Component Analysis(PCA). The algorithm involves comparison of trained samples with the query images. If the trained samples matches with the query images, then the detected face is a “known face”. If there is a mismatch, then the detected face is a “new face”. PCA extracts the eigen values and eigen vectors for the given set of samples in the PCA-sub space and compares it with the new values during run time. Then the threshold value is calculated from these eigen faces. If the new face value is less than the threshold, then it is a “known face”. Otherwise, he/she must be a new user. This algorithm is implemented using python on a Raspberry pi module in which the raspberry camera is used to extract the face values. The Face values are displayed on a 659M10 LCD so that it could act as an alarm in door security system.

Keywords: Arduino, Touchless Interface, Micro Controller, Capacitive Sensing

PDF | DOI: 10.17148/IJARCCE.2019.81012

Open chat
Chat with IJARCCE