Lightness enhancement method for low illumination night-time image.

Abstract

Computer vision applications under low illumination, such as night-time images, possess a difficult challenge. Night-time images suffer from the loss of visibility due to inadequate illumination, which affects the performance of outdoor computer vision systems. Thus, to improve the reliability of the outdoor computer vision systems during night-time, it is essential to enhance low illumination night-time images. In this paper, we propose a lightness enhancement method that removes the night-time color cast. The proposed method compensates for the red, green, and blue color channels of the night-time image by adding a fraction of each channel to itself. Then, it averages out the compensated color channel values with the compensated mean values. It effectively removes the night-time color cast and increases the image visibility. Experiments are conducted to evaluate the effectiveness of the proposed method, and the results are compared with two night-time image enhancement methods using objective and subjective evaluations. The results show that the proposed method produced enhanced images with better image visibility than the other two methods.