Position determination and face detection using image processing techniques and svm classifier

Abstract

In this paper, an improved algorithm for detecting the position of a person in controlled environments using the face detection algorithm is proposed. This algorithm ingeniously combines different face detection, occlusion detection algorithms and SVM classifier. A class room environment with thirty students is used along with some constraints such as position of the camera being fixed in a way that covers all the students, the static student's position and the class environment with the fixed lighting conditions. The students are treated as classes in this technique. For every class, a set of 6 attributes are derived and updated in a database. The image is given as an input to the face detection algorithm to detect some of the faces. Some faces are not detected because of occlusion, so an occlusion detection technique is implemented to detect all the occluded faces. In the training phase, a set of four images with the entire thirty students taken in four different days is used. Therefore a database of total 120 set of records with 6 attributes is used. © 2012 Springer-Verlag GmbH Berlin Heidelberg.