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

Image processing method and device as well as equipment

An image processing, image technology, applied in the field of image processing

Active Publication Date: 2018-01-05
NEW H3C TECH CO LTD
View PDF5 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] And there is usually redundant information between different images. For example, if the user collects multiple images with similar backgrounds, there is redundant information in these multiple images. When using existing solutions to compress these multiple images, it cannot Compression based on redundant information between images

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 processing method and device as well as equipment
  • Image processing method and device as well as equipment
  • Image processing method and device as well as equipment

Examples

Experimental program
Comparison scheme
Effect test

Embodiment approach

[0052] S102-S103 can be understood as a part of the neural network training process. As an implementation manner, the training process may specifically include:

[0053] 11. Preset the structure of the neural network: the structure includes a compressed neural network and a decompressed neural network, the two structures are symmetrical, and the connected parts ( figure 2 The central axis) is the axis of symmetry between the two.

[0054] As an implementation manner, the number of neurons in the input layer of the compressed neural network and the number of neurons in the output layer of the decompressed neural network=length*width*number of image channels in the above scaling parameters.

[0055] Continuing the above example, assuming that the image data contains the value of L*W*3 elements, the neuron data of the input layer of the initial neural network can be set to L*W*3, and each element corresponds to a neuron, correspondingly Symmetrically, the neuron data of the ou...

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 embodiment of the invention discloses an image processing method and device as well as equipment. The image processing method comprises the steps of training a compression neural network and a decompression neural network by utilizing multiple image data to obtain a compression calculation parameter and a decompression calculation parameter; and setting the calculation parameter of the compression neural network as the compression calculation parameter for compression of the image data so as to obtain compressed data of the image data, wherein a generated image compression package comprises compressed data of at least a portion of the image data, construction information of the first decompression neural network and a first decompression calculation parameter. Obviously, the compression calculation parameter in the scheme is the shared parameter of multiple image data, so that the compression calculation parameter is obtained by training redundant information among the image data.Therefore, each portion of the image data is compressed based on the redundant information among the image data, and compression based on the redundant information among the images is realized.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular to an image processing method, device and equipment. Background technique [0002] In order to reduce resources occupied by image data, such as network resources or storage resources, it is usually necessary to compress the image data. Generally speaking, there is redundant information in image data, such as spatial correlation redundancy, spectrum correlation redundancy, structure correlation redundancy, etc. Image compression is to remove part of the redundant information. [0003] Most of the existing image compression schemes are based on redundant information in a single image. For example, using DCT (Discrete Cosine Transform, Discrete Cosine Transform) coding compression technology, through frequency domain transformation, the original image data is converted into DC components and The coefficients of a small number of low-frequency AC components are used to...

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): H04N1/41H04N1/64
Inventor 梁鹏
Owner NEW H3C TECH CO LTD