A More Efficient Face Recognition Framework Based on Illumination Compensation, Kernel PCA and SVM

For which low frequency discrete cosine transform (DCT) coefficients retransforming based on contrast limiting adaptive histogram equalization (CLAHE) is proposed. Firstly, original images are divided into several non-overlapping blocks and CLAHE is used to do local contrast stretching so as to reduce noise. Then, illustration variation of face image is removed by reducing suit numbers of low frequency DCT coefficients. Finally, kernel principle component analysis is used to extract features and support vector machine is used to finish classification and recognition. Many experiments are carried out with the well-known databases like the Extended YaleB and FERET. Illustrative examples have been listed and the results compared to other advanced algorithms.