Image Gradient Regression Approach for Face Recognition

This paper represents a robust face recognition system. The proposed face recognition system consist is useful in forming an effective face recognition. Proposed method overcomes the problem of multi collinearity and also give good results in case of small sample size (SSS) problem . In Image Gradient Regression Approach (IGRA), first in preprocessing stage image is transformed by normalizing gradient of the image. Then image is converted into zero average data matrix. After preprocessing ridge regression is executed on projected data and identity is determined by calculating maximum correlation coefficient. Experiments carried out on faces94 and Pie facial databases reveal that the proposed framework outperforms the state-of-the-art methods and demonstrates promising abilities against face recognition.