A Simple Illumination Normalization Algorithm for Face Recognition

摘要: Most of the FR (face recognition) systems suffer from sensitivity to variations in illumination. For better performance system needs more training samples acquired under variable lightings but it is not practical real world. We introduce a novel pre-processing method, which makes illumination-normalized face image for recognition. The proposed ICR (Illumination Compensation based on Multiple Regression Model), find plane that best fits intensity distribution using multiple regression model, then use this normalize image. advantages our method are simple and practical. planar approximation mathematically defined by linear model. provide experimental results demonstrate public databases database. experiments show significant improvement recognition rate.