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

Image compression system, decompression system, training method and device, and display device

An image compression and decompression technology, applied in image communication, neural learning methods, biological neural network models, etc., can solve the problems of cumbersome and error-prone manual setting process

Inactive Publication Date: 2016-05-25
BOE TECH GRP CO LTD
View PDF9 Cites 36 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Generally, the filter parameters of each filter unit in such filters p and u are artificially set, and the weight values ​​set in this way are difficult to make the corresponding encoder obtain the best or near-optimal compression rate, and the artificial The setting process is extremely cumbersome and error-prone

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
  • Image compression system, decompression system, training method and device, and display device
  • Image compression system, decompression system, training method and device, and display device
  • Image compression system, decompression system, training method and device, and display device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0072] The structure of the image compression system provided by Embodiment 1 of the present invention can refer to figure 2 , including: splitting unit DM, image compression unit C, and has an input interface INPUT and four output interfaces.

[0073] The splitting unit DM is connected to the input interface INPUT, and is used to split each original image input to the input interface INPUT into four sub-images, for example, the first sub-image UL, the second sub-image UR, and the third sub-image BL respectively and the fourth sub-image BR; and for outputting the four sub-images to the image compression unit C through the first output terminal to the fourth output terminal for compression.

[0074] The image compression unit C includes: a first convolutional neural network module P, a difference acquisition module Y, a second convolutional neural network module U, and an image superposition module Z;

[0075] The first convolutional neural network module P is located between...

Embodiment 2

[0096] join Figure 4 , and in Example 1 figure 2 The difference is that the image compression system provided by the implementation of the present invention is a two-level image compression system, including two-level image compression units C1 and C2 and four output interfaces; each level of image compression unit also includes the first convolution The neural network module, the second convolutional neural network module, the difference acquisition module, and the superposition module; for the convenience of distinction, in Figure 4 In , the first convolutional neural network module in the first-level image compression unit C1 is represented as P2, the second convolutional neural network module is represented as U2, the difference acquisition module is represented as Y2, and the image superposition module is represented as Z2; and The first convolutional neural network module in the second-level image compression unit C2 is represented as P1, the second convolutional neu...

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 discloses an image compression system, a decompression system, a training method and device, and a display device. In the image compression system, a convolution neutral network module is utilized to complete updating and predicting processes, so that a corresponding image compression system has a better compression ratio when weight of each filter unit in the convolution neutral network module is trained. Therefore, the difficulty of setting filtering parameters of an image compression unit and an image decompression unit is reduced.

Description

technical field [0001] The present invention relates to the field of display technology, in particular to an image compression system, a decompression system, a training method and device, and a display device. Background technique [0002] Wavelet transform is a method of multi-resolution image conversion, which is often used in image compression. The application of wavelet transform includes transform transcoding in JPEG2000 standard. The purpose of wavelet transform is to represent the entire original image by a part of the entire image, and the original image can be obtained by using the low-resolution image (a part of the original image) and some difference features required to restore the entire original image. Lifting is an efficient implementation of wavelet transforms and is also a flexible tool for constructing wavelets. figure 1 A typical structure of 1D data is shown in . The left corresponds to the encoder. The encoder uses a prediction filter p and an updat...

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
Patent Type & Authority Applications(China)
IPC IPC(8): H04N19/635H04N19/117H04N19/42H04N19/48
CPCH04N19/117H04N19/42H04N19/48H04N19/635H04N19/90G06N3/08G06N3/045H04N19/50
Inventor 那彦波张丽杰李晓宇何建民
Owner BOE TECH GRP CO LTD
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