Gradient-based local descriptor and centroid neural network for face recognition
ISNN'10 Proceedings of the 7th international conference on Advances in Neural Networks - Volume Part II
| Hi-index | 0.00 |
The rising recognition rate is prominent in the rapid development of face recognition. However, much information shows that illumination, expression, and disguise are still a challenge needed to be conquered. In this paper, we propose an algorithm which could decrease the effect of illumination. We use Sobel as preprocess and median filtering to promote the accuracy of covariance matrix. After that, eigen value and vectors are calculated. Finally, we use eigen vectors to project the Sobel face into Eigen face. We use ORL and YALE databases to test the algorithm. The experiment show that our algorithm has better result than PCA, 2DPCA, and LDA, especially in YALE database.