Facial Recognition based Attendance

This paper shows an approach for face detection which can process images and attaining high detection rates. that is accomplished via a three-key concept - integral image, AdaBoost and Cascade. An integral image is a new image, introduced to permit the features to be computed. The AdaBoost (Adaptive Boosting), a statistical classification meta-algorithm, picks a certain number of distinguishable features from a set and returns classifiers that efficiently detect those features. The Cascade is a technique where an image is traversed pixel by pixel. Complex classifiers are combined during this traversal ignoring the image background and spending more computation on the region that appears to be a face. It can be viewed as a focusing mechanism which targets specific objects and ensures that the discarded regions aren't probable to comprise of the object of interest. This approach can be used to record attendance. This saves time and is a cheat free method to record attendance.