Informatics Publishing Limited

Uma Shankar Kurmi ¹, Dheeraj Agrawal ¹, R. K. Baghel ¹

Affiliations
1 Dept. of Electronics & Communication, Maulana Azad National Institute of Technology, Bhopal, IN

Abstract

At present face recognition has wide area of applications such as security, law enforcement. Imaging conditions, Orientation, Pose and presence of occlusion are huge problems associated with face recognition. The performance of face recognition systems decreases due to these problems. Discriminant Analysis (LDA) or Principal Components Analysis (PCA) is used to get better recognition results. Human face contains relevant information that can extracted from face model developed by PCA technique. Principal Components Analysis method uses eigenface approach to describe face image variation. A face recognition technique that is robust to all situations is not available. Some techniques are better in case of illumination, some for pose problem and some for occlusion problem. This paper presents some algorithms for face recognition.

Keywords

Eigenfaces, Recognition, PCA, LDA.

Full Text

   

Uma Shankar Kurmi ¹, Dheeraj Agrawal ¹, R. K. Baghel ¹

Affiliations
1 Dept. of Electronics & Communication, Maulana Azad National Institute of Technology, Bhopal, IN

Abstract

This paper presents a face completion and recognition algorithm under partial occlusion. Component-based algorithm utilized in the proposed method. Individual components are detected using Haar-like features and wild card estimation is used for face completion task. The system performs better results under partial occlusion conditions in comparison to existing methods. AR face database is used for experiments.

Keywords

Face Recognition, Face Completion, Occlusion, Haar-Features, Topology, Components.

Full Text