Image Enhancement Using Optimized Generative Adversarial Networks

Poor quality and degradation in images are frequently experienced in applications such as surveillance, medical imaging and entertainment. As a result, there is a growing demand for effective image enhancement techniques in various fields which utilizes image processing. The aim of this research is to meet this need. The suggested method addresses image restoration problems by combining a CNN model-VGG16 in this case with Generative Adversarial Networks (GANs). A CNN extracts feature loss from a low-quality input image. Then a Generative Adversarial Network creates a high-resolution image based on the feature loss it has extracted from the input image. Also a new loss function is suggested that combines the features, edges, and reconstruction losses into one. Using deep learning techniques, the combination of CNNs and GANs produces enhanced high-quality images from low-resolution images with finer edges.