Self-quotient-referential bilateral filter for images under varying lighting conditions

This paper presents edge-preserving smoothing by self-quotient-referential bilateral filter for images under varying lighting conditions. Edge preserving smoothing is a fundamental feature extraction from the image for industrial applications. Bilateral filter is a promising nonlinear filter, which can smooth the image while preserving edge information. However, when we handle the images under light variation, the contrast of edge part is high in high contrast area, while the contrast of edge part is low in low contrast area. As bilateral filter basically uses the information of local intensity to preserve edge information, it is difficult to handle these types of images. It cannot preserve the edge information around low contrast area. Our method solves this problem by referring the self-quotient image instead of the original image in bilateral filter. To show the effectiveness of our method, we conducted some comparison experiments.