The Atrous CNN Method with Short Computation Time for Super-Resolution

In this paper, we proposed the atrous CNN method with the shorter computation time while maintaining a high quality of image compatible to the VDSR method. We found that the computation was faster than VDSR through making experiments 30 times. To verify whether image quality for the proposed method is significant statistically or not, we evaluated the quality of the reconstruction image for 100 images. Through the ANOVA analysis, we found that there was no significant difference between methods in the view of the PSNR value and was a significant difference between methods in the view of the SSIM value. As the result of post-hoc analysis, there were two groups; one was the proposed method and the VDSR method. The other was the SRCNN method. In conclusion, the proposed method met our goal of maintaining compatible image quality and reducing computation time compared to VDSR method.

[1]  Narendra Ahuja,et al.  Deep Laplacian Pyramid Networks for Fast and Accurate Super-Resolution , 2017, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[2]  Michal Irani,et al.  Super-resolution from a single image , 2009, 2009 IEEE 12th International Conference on Computer Vision.

[3]  Wei Wang,et al.  Deep Learning for Single Image Super-Resolution: A Brief Review , 2018, IEEE Transactions on Multimedia.

[4]  George Papandreou,et al.  Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation , 2018, ECCV.

[5]  Yusra A. Y. Al-Najjar,et al.  Comparison of Image Quality Assessment: PSNR, HVS, SSIM, UIQI , 2012 .

[6]  Kyoung Mu Lee,et al.  Accurate Image Super-Resolution Using Very Deep Convolutional Networks , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[7]  Andrew McCallum,et al.  Fast and Accurate Entity Recognition with Iterated Dilated Convolutions , 2017, EMNLP.

[8]  Xiaoou Tang,et al.  Image Super-Resolution Using Deep Convolutional Networks , 2014, IEEE Transactions on Pattern Analysis and Machine Intelligence.