Unlock instant, AI-driven research and patent intelligence for your innovation.

Combining dct domain and pixel domain learning method to decompress jpeg compressed image

A technology for compressing images and pixel domains, applied in the field of JPEG compressed image decompression effect and quality improvement of compressed images

Active Publication Date: 2022-03-04
SICHUAN UNIV
View PDF7 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the current image decompression effect algorithm based on convolutional neural network still has room for further improvement in terms of prediction performance and network structure effectiveness.

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
  • Combining dct domain and pixel domain learning method to decompress jpeg compressed image
  • Combining dct domain and pixel domain learning method to decompress jpeg compressed image
  • Combining dct domain and pixel domain learning method to decompress jpeg compressed image

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0018] The present invention will be further described below in conjunction with accompanying drawing:

[0019] figure 1 Among them, the JPEG compressed image decompression effect method combined with DCT domain and pixel domain learning can be divided into the following five steps:

[0020] (1) For JPEG compressed images, respectively construct DCT domain network and pixel domain network structure models based on convolutional neural network;

[0021] (2) Utilize the training image set to train the two convolutional neural network models constructed in step (1) respectively;

[0022] (3) Utilize two convolutional neural networks that train in step (2) to obtain, predict output to JPEG compressed image respectively;

[0023] (4) Using a weighted average method, the two prediction results in step (3) are fused;

[0024] (5) Reconstitute the fused image tensor into an image to obtain the final decompression effect processing result.

[0025] Specifically, in the step (1), th...

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 a JPEG compressed image decompression effect method combining DCT domain and pixel domain learning. It mainly includes the following steps: for JPEG compressed images, construct DCT domain network and pixel domain network structure models based on convolutional neural network respectively; use the training image set to train the constructed two convolutional neural network models respectively; A convolutional neural network is used to predict and output JPEG compressed images respectively; the prediction results of the two networks are fused by means of weighted average; the fused image tensors are recombined into images to obtain the final decompression effect processing result . The method of the invention can reduce the compression effect existing in the JPEG compressed image, and restore the detail information lost in the compression process of the image to a certain extent. The method of the invention can be applied to the fields of image and video compression, digital multimedia communication and the like.

Description

technical field [0001] The invention relates to a technology for improving the quality of compressed images, in particular to a method for decompressing JPEG compressed images combined with DCT domain and pixel domain learning, and belongs to the field of digital image processing. Background technique [0002] Image compression is an image processing technology that reduces the amount of data by reducing the redundancy between original image data, which can save storage space and bandwidth resources to a certain extent. However, when the coding bits are limited, that is, the compression factor is high, there will be obvious distortion and compression effect in the compressed image, which seriously reduces the subjective and objective quality of the compressed image and limits its further application. [0003] The essence of the distortion produced by the JPEG compression method is that the DCT coefficients are quantized before entropy coding, which introduces rounding errors...

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 Patents(China)
IPC IPC(8): H04N19/625H04N19/61H04N19/70G06T9/00
CPCH04N19/625H04N19/61H04N19/70G06T9/002
Inventor 何小海李兴龙任超孙梦笛熊淑华普拉迪普卡恩滕奇志
Owner SICHUAN UNIV