10 May 2018 Image deblocking via joint domain learning
Author Affiliations +
Abstract
Image deblocking is a postprocessing method that aims to suppress the compression artifacts without changing existing JPEG coding standard. We propose an image deblocking method, which is based on deep convolutional neural networks. The proposed method takes full advantage of the characteristics of wavelet domain and pixel domain to restore the high-frequency information of compressed images and maintain low-frequency information, respectively. In addition, a fusion layer is employed to fuse the merits of two domains. Extensive experiments demonstrate that the proposed method outperforms the state-of-the-art deblocking methods in both subjective vision and objective evaluation.
© 2018 SPIE and IS&T
Wenshu Zhan, Xiaohai He, Shuhua Xiong, Chao Ren, Honggang Chen, "Image deblocking via joint domain learning," Journal of Electronic Imaging 27(3), 033006 (10 May 2018). https://doi.org/10.1117/1.JEI.27.3.033006 Submission: Received 15 January 2018; Accepted 17 April 2018
Submission: Received 15 January 2018; Accepted 17 April 2018
JOURNAL ARTICLE
12 PAGES


SHARE
Back to Top