A New Face Recognition Algorithm Based on Haar-Like Features and SRC with Gentle Adaboost
- DOI
- 10.2991/aiie-15.2015.4How to use a DOI?
- Keywords
- Haar-Like features; Gentle Adaboost; sparse representation; face recognition
- Abstract
In this paper, a new face recognition algorithm based on Haar-Like features and Gentle Adaboost (GA) feature selection via sparse representation was proposed. Firstly, all the images including face images and non-face images were normalized to size 20×20, then Haar-Like features were extracted from the images after size normalization. The number of the extracted features is more than 125,199. Then, Gentle Adaboost algorithm was exploited to feature selection that reduced the feature dimension and retained the most effective features for face recognition. Finally, the samples were classified and identified for face recognition via sparse representation classification (SRC) algorithm. Comparing it with SRC, Nearest Neighbor (NN), Nearest Subspace (NS), Support Vector Machine (SVM) algorithms, the experiment results on AR database demonstrate that the new proposed algorithm can achieve a higher recognition rate than these traditional algorithms. Even though the increase of the dimension, the new proposed algorithm always got higher recognition rate than SRC and other algorithms, which proved the new algorithm can get better identification effects and stronger stability in the low dimension feature space.
- Copyright
- © 2015, the Authors. Published by Atlantis Press.
- Open Access
- This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).
Cite this article
TY - CONF AU - G.M. Zhu AU - Z.L. Ying AU - L.W. Huang PY - 2015/07 DA - 2015/07 TI - A New Face Recognition Algorithm Based on Haar-Like Features and SRC with Gentle Adaboost BT - Proceedings of the 2015 International Conference on Artificial Intelligence and Industrial Engineering PB - Atlantis Press SP - 10 EP - 12 SN - 1951-6851 UR - https://doi.org/10.2991/aiie-15.2015.4 DO - 10.2991/aiie-15.2015.4 ID - Zhu2015/07 ER -