Facial Detection Attendance System using LBPH and KNN

Present procedure for marking attendance is exhausting and prolonged so this paper is consequently put forward to challenge all these complications. This paper hence evolve a model to distinguish each personality's face from a seized image using a set of conditions i.e. LOCAL BINARY PATTERN HISTOGRAM algorithm to track the student attendance. The overall working of this local binary pattern histogram algorithm was, first the image is divided into m*m grids. For each grid histogram is calculated in order to easily recognize the spatial features. After calculating binary pattern histogram for each cell. The results were coupled to obtain the final feature vector. This final vector is compared with vectors in the training data set using K-Nearest Neighbor's algorithm. By this algorithm the value which is closest to our final vector is obtained as a result of classification. After receiving the name of the person, the attendance of the particular person is updated in the database. This proposed algorithm decreases the work load and records routine performance of maintaining each student and further makes it easy to note the attendance.