• Deepika Pal
• Prof. Sumit Sharma,

In this research work presented a Modified Very Deep Convolutional Networks (MVDCN) for single image super resolution (SISR). The proposed method based on modified CNN, in which apply different image feature for training also apply up-sampling as well as residual Images that is fundamental step of SISR, and dept of network is 20. For the improvement of presented method result apply fusion of two bicubic method attributes with Very Deep Convolutional Networks. The presented method shows better result in terms of two base parameters of proposed method that is PSNR and SSIM. There are different data set available in the for training and testing of presented method such Test data set Datasets ‘Set5’ [15] and ‘Set14’ [26] both are mainly used by different researcher, benchmark in other works data set ‘Urban100’, that’s very interesting as it contains many challenging images failed by many of the existing methods.The proposed method compare with different methods, they are Ground Truth data set image A+, RFL, SelfEx, and VDSR. The proposed and presented method “Modified Very Deep Convolutional Networks (MVDCN) for single image super resolution (SISR)” shows better result as compare to other previous methods.

Single Image Super Resolution (SISR), CNN, bicubic, Deep Convolutional Networks (DCN), ‘Set5’ ‘Set14’ and ‘Urban100’.

"A Single Image Super Resolution using Deep CNN for natural images", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.8, Issue 11, page no.a99-a105, November-2021, Available :http://www.jetir.org/papers/JETIR2111014.pdf

"A Single Image Super Resolution using Deep CNN for natural images", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.8, Issue 11, page no. ppa99-a105, November-2021, Available at : http://www.jetir.org/papers/JETIR2111014.pdf