Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Implementation method and system for video coding and decoding in-loop filter based on a convolutional neural network

A convolutional neural network and video encoding and decoding technology, applied in the field of filtering implementation methods and systems in video encoding and decoding loops, can solve problems such as limited encoding efficiency and limited compression ratio, achieve good compression ratio and encoding efficiency, and have robustness. Sticky, widely applicable effect

Active Publication Date: 2018-06-08
SHANGHAI JIAO TONG UNIV
View PDF6 Cites 13 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, due to the limitations of the intra-frame reference mode, the compression ratio is very limited, and the quality improvement of the reconstructed frame by the out-of-loop filter will not produce gains in the subsequent encoding process. very limited improvement

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Implementation method and system for video coding and decoding in-loop filter based on a convolutional neural network
  • Implementation method and system for video coding and decoding in-loop filter based on a convolutional neural network
  • Implementation method and system for video coding and decoding in-loop filter based on a convolutional neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0018] Such as figure 1 As shown, this embodiment relates to a video codec in-loop filter based on convolutional neural network.

[0019] The implementation method under the reference software HM-16.0 of h.265 / HEVC video coding standard, the specific steps are as follows:

[0020] Step 1) Use the video codec software HM to code and decode a series of videos, and finally get the reconstructed video after decoding. After preprocessing, the training data is obtained, and the preset convolutional neural network is trained to optimize its model parameters. In this embodiment, a basic network with only images as input and no additional branches is used.

[0021] The pre-processing refers to dividing each decoded video frame, the original video frame, and the division graph of the video frame, using only the Y channel of the image, and dividing them into 64x64 disjoint subgraphs. The division graph is as follows figure 2 Shown.

[0022] Such as image 3 As shown, the convolutional neural ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention provides a video coding and decoding in-loop filter based on a convolutional neural network and an implementation method thereof. The implementation method comprises the following steps:taking a video obtained by coding and decoding through a video coding and decoding algorithm as training data, training the convolution neural network by using a supervised learning method to obtaina pre-training model; dividing each reconstructed frame into a plurality of sub-graphs in a video coding and decoding loop; adopting the pre-training mode and taking each sub-graph as the input, and outputting an image with the same size as an input image; and selectively using an output image to update an original image according to the condition that whether the quality of the output image is improved or not. According to the implementation method of the video coding and decoding in-loop filter based on the convolutional neural network provided by the invention, the image quality of the reconstructed frame in the coding and decoding process can be improved, and a gain is provided for the subsequent coding process, so that the efficiency of the coding algorithm is improved finally.

Description

Technical field [0001] The invention relates to a technology in the field of digital image processing, in particular to a method and system for implementing in-loop filtering of video coding and decoding based on convolutional neural networks. Background technique [0002] The existing video coding and decoding algorithms have two main components: an encoder and a decoder: the encoder is responsible for encoding the input video into a bitstream according to the video coding standard, and the decoder is responsible for decoding the bitstream to obtain the final decoded video. In the encoding and decoding process, the encoding and decoding loop filter and the loop filter are used to improve the image quality of the reconstructed video frame. [0003] The existing in-loop filter uses Deblock Filter and Sample Adaptive offset Filter as the loop filter, that is, in the encoding process, the reconstructed image is filtered to improve the image quality. The characteristic of the in-loop ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): H04N19/117H04N19/80H04N19/154H04N19/172
CPCH04N19/117H04N19/154H04N19/172H04N19/80
Inventor 林巍峣何晓艺
Owner SHANGHAI JIAO TONG UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com