An Image Preprocessing Algorithm for Illumination Invariant Face Recognition

Face recognition algorithms have to deal with significant amounts of illumination variations between gallery and probe images. State-of-the-art commercial face recognition algorithms still struggle with this problem. We propose a new image preprocessing algorithm that compensates for illumination variations in images. From a single brightness image the algorithm first estimates the illumination field and then compensates for it to mostly recover the scene reflectance. Unlike previously proposed approaches for illumination compensation, our algorithm does not require any training steps, knowledge of 3D face models or reflective surface models. We apply the algorithm to face images prior to recognition. We demonstrate large performance improvements with several standard face recognition algorithms across multiple, publicly available face databases.