A Neural Network-Based Algorithm for Weak Signal Enhancement in Low Illumination Images

There is noise interference in low-illumination images, which makes it difficult to extract weak signals. For this reason, paper proposes a image signal enhancement algorithm based on neural network. Multi-scale normalization performed low-light and multi-scale Retinex used enhance signals images. On basis, the GAN artificial network detect of image, completed residual network, self-encoding parameters depth are generated, result output. The experimental results show that method has better effect images denoising effect. When scale value large, low-contrast area saturated