NVGAN: A GAN-Based Night Vision Image Enhancement Model

Night vision image enhancement is the process of improving the perception and clarity of an image by processing it at night or under low light conditions. To achieve this effect, the mainstream approach is to adopt traditional image processing methods or to adopt different learning strategies based on deep learning. In this paper, we propose a GAN-based night vision image enhancement model, NVGAN, which consists of two networks: a generative network and a discriminative network. The generative network is used for illumination enhancement and deep and shallow feature extraction and enhancement for night vision images, and the discriminative network is selected as DnCNN for suppressing the effect of noise on night vision images. Meanwhile, the loss function is further improved to obtain a better performance for the GAN network in order to obtain a better enhancement effect.