FAST SUPER RESOLUTION IMAGE USING CONVOLUTION NETWORK
Keywords:
Total variation method, Gaussian noise PSSR, Up-sampling TV, CNN, neural networkAbstract
Aim of Super resolution is to generate high-resolution image from single or multiple low resolution of the
same picture or image. Super resolution method which based on Total Variation regularization and total variation up
sampling with Gaussian noise PSSR which provide better resolution image with preserving only texture components.
More processing time due to the calculation of total variation in the input Image. To overcome problem of existing work,
Novel approach of convolution neural network (CNN) which is consists of three layers namely convolution layer, maxpooling layer and reconstruction layer. This approach will try to reduce the processing time as well as increasing the
PSNR ratio, RMSE And SSIM. The experimental results show that the proposed method can perform better than the
wavelet transform based method in some situations.
