Journals Proceedings

International Journal of Image Processing Techniques

Robust Content-based Image Retrieval System for Face Recognition

Author(s) : SUHAIB KAMAL , ZHIYONG LI

Abstract

many types of biometric systems are used for different purposes. Face recognition is considered the most effective biometric system due to the fact that it requires the least amount of human interaction. This attracted many researchers to pay a lot of attention to the field of face recognition and its applications. One of the approaches used in face recognition systems is using Content-Based Image Retrieval (CBIR). As its name implies, CBIR uses the actual visual contents of an image to retrieve similar images from a database or a set of images. The visual contents of images are described with features obtained using various feature extraction techniques. This research proposes a new robust face recognition system based on CBIR. The efficiency of two types of mathematical features is evaluated and compared in terms of accuracy, required CPU processing time and robustness against different challenges such as changes in scaling and lighting conditions in order to use the best features as new face recognition features. The mathematical investigated features are: Harris-Stephens features and Minimum Eigenfeatures. The proposed system combines several methods to implement a fast and accurate CBIR system for face recognition. The proposed system is also compared with other modern face recognition systems using two published face images databases

No fo Author(s) : 2
Page(s) : 17 - 24
Electronic ISSN : 2372-3998
Volume 2 : Issue 2
Views : 363   |   Download(s) : 127