A novel approach of low-light image used for face recognition

Abstract

Illumination variation is a challenge of face recognition, especially in low light environments. In order to overcome the influence of low illumination image, this paper proposed a face recognition method of face image preprocessing before recognition, with illumination- reflection model of homomorphic filtering and image multiplication method of image preprocessing for getting the enhancement of human face recognition rate. Homomorphic filtering can reduce low frequency noise in the frequency domain and enhance image details at same time reduce the high frequency noise of the image. By changing the image multiplication coefficient can change the brightness of the image. The YALE database results show that this approach can be a very good solution to the problem of face recognition under low light conditions. Human face recognition rate can increase 10%. The experiment proved that the pretreatment method can well solve the problem of face recognition under low light.