Analysis of Face Pattern Detection Using the Haar-Like Feature Method

Abstract

Student attendance is one of the important factors in the processing of discipline, obligations and obedience of students following the lecture process. Student attendance recordings are done manually by using signatures, manual attendance recording can be a barrier to monitoring discipline, obedience of students in terms of punctuality incoming students. Manually attendance records can be replaced by computerized attendance records using the biometric technology identification process, to identify student face patterns using bilateral filter methods, canny edge detection, Haar-like features, integral images, adaboost cascade classifier. From the test results, the success rate of student attendance recording based on the facial pattern of each student is 70.43%.