$$l_{2}$$ Norm Prior-Based Modified Bright Channel for Low-Illumination Images

Low-illumination image enhancement problem is a very challenging problem in many computer vision applications and, when it comes to nighttime low-illumination images, it becomes more challenging because the depth information of the low-illumination image is not known. Recently, bright channel prior-based methods are used to enhance the overall illumination of the image. The bright channel prior is based on statistical observation on the low-illumination image containing some regions with bright intensity pixels. In this paper, we propose an improved \(l_{2}\) norm-based prior bright channel to enhance the overall illumination of the image by maintaining the image contrast. This new generated bright channel is free from the block effect, which makes our method more robust than other methods. The experimental results show the effectiveness of the proposed method on the low-illumination images as well as on the nighttime low-illumination images.