Face Identification Using Histogram based Approach With a Probabilistic ClassifierIn Spatial Domain

Authors

  • Pankaj Prusty, Bibhu Prasad Mohanty

Abstract

Face identifiaction is a well known and widely used method in computer vision and also in many security related biometeric applications. The major problem that the face detection algorithm encounters is thge variation in position, illumination due to different conditions and many more. In this paper we have focused on the fact that the test input image must correspond to a predefined image in the database. The major feature used here is Histogram of oriented Gradients. The feature vector is used to train the probabilistic classifier ,so that the classifier can easily distinguish the data set image and non data set image. The method also verifies the correct image from a data set on ce found to belong the data set. At the end the performance of the algorithm is verified with the help of correct and miss classification.

Keywords: Histogram of oriented gradients, probabilistic classifier, miss classification

Downloads

Published

2020-05-18

Issue

Section

Articles