Title:
METHOD AND APPARATUS FOR SUPER-RESOLUTION RECONSTRUCTION OF MULTI-SCALE DILATED CONVOLUTION NEURAL NETWORK
Document Type and Number:
WIPO Patent Application WO/2020/056791
Kind Code:
A1
Abstract:
Disclosed in the present invention are a method and apparatus for super-resolution reconstruction of a multi-scale dilated convolution neural network. By constructing a dilated convolution super-resolution network, the problem in the prior art of a small receptive field is effectively solved, a network receptive field is increased, and moreover, in order to make the dilated convolution super-resolution network suitable for super-resolution reconstruction of different scales, multi-scale data enhancement is performed on an original image, allowing the dilated convolution super-resolution network to process images with different multiples of size, so that the dilated convolution super-resolution network can generalize the multi-scale image super-resolution.
Inventors:
XU YING (CN)
YING ZILU (CN)
SHANG LIJUAN (CN)
ZHAI YIKUI (CN)
WANG TIANLEI (CN)
GAN JUNYING (CN)
ZENG JUNYING (CN)
QIN CHUANBO (CN)
CAO HE (CN)
DENG WENBO (CN)
YING ZILU (CN)
SHANG LIJUAN (CN)
ZHAI YIKUI (CN)
WANG TIANLEI (CN)
GAN JUNYING (CN)
ZENG JUNYING (CN)
QIN CHUANBO (CN)
CAO HE (CN)
DENG WENBO (CN)
Application Number:
PCT/CN2018/107916
Publication Date:
March 26, 2020
Filing Date:
September 27, 2018
Export Citation:
Assignee:
UNIV WUYI (CN)
International Classes:
G06T3/40
Foreign References:
| CN108537731A | 2018-09-14 | |||
| CN108428212A | 2018-08-21 | |||
| CN107392852A | 2017-11-24 | |||
| CN107464216A | 2017-12-12 |
Attorney, Agent or Firm:
JIAQUAN IP LAW FIRM (CN)
Download PDF: