An Approach for Face Detection Using HAAR Features for Automatic Attendance System

Human face detection has a wide application in biometric and security fields. It finds its application in photography, video surveillance and in many other areas. This paper proposes a face detection approach using Viola Jones algorithm with a Haar features converted to integral image to decrease the computational complexity. AdaBoost algorithm is used for feature selection and attentional cascade for fast rejection of non-face windows. This approach performs fifteen times faster than other face detection algorithms. This method is implemented in college for automatic class attendance system, which takes real time image as input and detects faces crops the face region and display along with their count, which helps in taking automatic attendance. With a high accuracy, the detected faces along with the count are updated. An improved result for face detection and accuracy is obtained using this algorithm. Keywordsface detection, haar features, Integral image, Adaboost algorithm