An Improved Convolutional Neural Network and Its Applications to Image Super Resolution

This paper studies an improved convolutional neural network application in the image super-resolution problem. On the traditional super-resolution model, we improve the model framework, the upsampling method and the network design. And through the experiment of traditional super-resolution convolution neural network with improved super-resolution convolution neural network and compare the effect of the reconstruction image, it is concluded that the improved super-resolution network model compared with super-resolution in higher PSNR and better visual effect, stronger image resolution.