Melika A’la - Department of Mathematics and Computer Science ,Amirkabir University of Technology;
Mohammad Ebrahim Shiri - Department of Mathematics and Computer Science, Amirkabir University of Technology;
Hedieh Sajedi - School of Mathematics, Statistics and Computer Science, College of Science, University of Tehran;

Recent studies have shown that the performance of single image super-resolution methods can be significantly boosted by using deep convolutional neural networks. Lately a method has been proposed which directly learns an end-to-end mapping between the low/high-resolution images. This mapping is done by a deep Convolutional Neural Network (CNN) that takes the lowresolution image as the input and outputs the high-resolution one. In this paper, we enhance this method by using the canny edge detection method. A combination of Deep CNNs and Skip connection layers are used as a feature extractor for image features, Moreover, deconvolution layers are integrated into the network to learn the up sampling filters and to speed up the reconstruction process. The results of the proposed method shows that considering the result of Canny edge detection in producing the high resolution image, results in better output.

Deep Learning, Image Super Resolution, Deep CNN, Residual Net, Skip Connection, Network in Network.