Comparative analysis of face recognition techniques with illumination variation
K. C. Jondhale, Laxman Madhavrao Waghmare · IOP Conference Series Materials Science and Engineering · 2010
Illumination variation is one of the major challenges in the face recognition. To deal with this problem, this paper presents comparative analysis of three different techniques. : First, the DCT is employed to compensate for illumination variations in the logarithm domain. Since illumination variation lies mainly in the low frequency band, an appropriate number of DCT coefficients are truncated to reduce the variations under different lighting conditions. The nearest neighbor classifier based on Euclidean distance is employed for classification. Second, the performance of PCA is checked on normalized image. PCA is a technique used to reduce multidimensional data sets to a lower dimension for analysis. Third, LDA based methods gives a satisfactory result under controlled lighting condition. But its performance under large illumination variation is not satisfactory. So, the performance of LDA is checked on normalized image. Experimental results on the Yale B and ORL database show that the proposed approach of application of PCA and LDA on normalized dataset improves the performance significantly for the face images with large illumination variations.