Convolution Neural Network Approach for Single Image Super Resolution

The goal of Super-Resolution (SR) is to generate a higher-resolution image from lower-resolution input images. High-resolution images offer more pixel density, thus capturing finer details of the original scene. Single Image Super-Resolution (SISR) seeks to restore a high-resolution image from a single low-resolution input, which is a significant challenge in computer vision. This process involves using the low-resolution image as the input and the high-resolution image as the reference, with the SR model producing the predicted high-resolution output. This paper proposes a neural network-based approach utilizing convolutional layers to improve Peak Signal-to-Noise Ratio (PSNR) and reduce processing time compared to traditional methods. The architecture consists of a convolutional layer, a max-pooling layer, and a reconstruction layer.