Research on Adaptive Face Recognition Algorithm Under Low Illumination Condition

Abstract

With the popularization of face recognition technology, the demand for the accuracy of face recognition has greatly increased. In the face image acquisition, the low illumination environment has a significant effect on the quality of human face images. Face images are susceptible to many factors such as image background, brightness, and image noise. This leads to many problems such as low detection rate of face recognition and the decline of recognition accuracy. In this paper, we take the face images under low illumination condition as a sample and present an adaptive algorithm based on the OTSU segmentation algorithm, which realizes the self-adaptation image acquisition under low illumination condition. By using the Adaboost classification detector to verify low-light face images before and after processing, the model we used in this paper successfully improved the accuracy of face detection.