Fusing Gabor and LBP feature sets for KNN and SRC-based face recognition

Citations Over TimeTop 10% of 2016 papers

Abstract

One of the major challenges in face recognition is that related to the differences in orientation or pose, the variations of illumination, the facial expressions, the occlusions and aging. In this paper, we propose an efficient method for face recognition in an uncontrolled environment where we fuse Gabor wavelets and Local Binary Patterns (LBP) in the feature extraction phase. Then, we apply the dimension reduction technique to reduce the pattern vectors. At last, we combine both K Nearest Neighbor (KNN) and Sparse Representation Classifier for face recognition phase. We evaluate our method on LFW database and we perform a comparative experimental study using several experiments. Obviously, the best result is obtained with a recognition rate = 94.16 per cent.

Related Papers

• → A Novel Crowd Density Estimation Technique using Local Binary Pattern and Gabor Features(2017)17 cited
• → Texture Classification Framework Using Gabor Filters and Local Binary Patterns(2018)2 cited
• → <title>Texture analysis using Gabor wavelets</title>(1996)15 cited
• A Study on the Extraction of the Texture Feature in the Process of Image Retriecal(2006)
• → Gabor and Log-Gabor Wavelet for Face Recognition(2010)