Face Recognition based on Modular 2DPCA and Contextual Constraints based Kernel Discriminant Analysis
Hua-Li Feng · Proceedings of the 2nd International Conference on Computer Science and Electronics Engineering (ICCSEE 2013) · 2013
In this paper, an improved face recognition algorithm is proposed based on the combination of modular 2DPCA and contextual constraints based kernel discriminant analysis (CCKDA) because of the disadvantages of CCLDA.CCLDA first transforms an image matrix to a vector which caused high dimensionality and computational complexity and not considers the local feature.While our method first extracts the local features with the original images which are divided into modular sub-images, then CSKDA is utilized, which incorporates the contextual information into kernel discriminant analysis.Experimental results obtained on ORL and XM2VTS databases show the effectiveness of the new method.