Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
VIDEO BLIND SUPER-RESOLUTION RECONSTRUCTION METHOD AND SYSTEM BASED ON SELF-SUPERVISED LEARNING
Document Type and Number:
WIPO Patent Application WO/2022/155990
Kind Code:
A1
Abstract:
Provided are a video blind super-resolution reconstruction method and system based on self-supervised learning, the method comprising: firstly, determining, by using a self-supervised learning method, a blur kernel estimation network, an optical flow estimation network, a feature extraction network and a potential high-resolution intermediate frame reconstruction network; estimating a blur kernel by using a first-resolution video sequence on the basis of the blur kernel estimation network; then, determining a deformation matrix on the basis of the optical flow estimation network and the first-resolution video sequence; then extracting, by using the feature extraction network, features of each video frame in the first-resolution video sequence, and aligning the features of each video frame according to the deformation matrix; constructing second-resolution intermediate video frames by using the potential high-resolution intermediate frame reconstruction network and the aligned features of each video frame; and finally determining a second-resolution video on the basis of the second-resolution intermediate video frames. In the present invention, by means of the self-supervision method, artifacts and incorrect structure information can be effectively ameliorated during high-resolution video reconstruction, thereby further improving the visual effect.

Inventors:
PAN JINSHAN (CN)
BAI HAORAN (CN)
TANG JINHUI (CN)
Application Number:
PCT/CN2021/074281
Publication Date:
July 28, 2022
Filing Date:
January 29, 2021
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
UNIV NANJING SCI & TECH (CN)
International Classes:
G06T3/40
Domestic Patent References:
WO2003041562A22003-05-22
Foreign References:
CN112218072A2021-01-12
CN105913394A2016-08-31
US20140354886A12014-12-04
Other References:
JINSHAN PAN; SONGSHENG CHENG; JIAWEI ZHANG; JINHUI TANG: "Deep Blind Video Super-resolution", ARXIV.ORG, CORNELL UNIVERSITY LIBRARY, 201 OLIN LIBRARY CORNELL UNIVERSITY ITHACA, NY 14853, 10 March 2020 (2020-03-10), 201 Olin Library Cornell University Ithaca, NY 14853 , XP081618462
"Master Thesis", 1 June 2020, HARBIN INSTITUTE OF TECHNOLOGY, CN, article BI MINMIN: "Research on Image Super Resolution Technology Based on Deep Learning", pages: 1 - 66, XP055952717
Attorney, Agent or Firm:
BEIJING GAOWO LAW FIRM (CN)
Download PDF: