Dictionary learning static image lossy compression method based on minimum quantization error criterion

A technology of quantization error and dictionary learning, which is applied in image coding, image data processing, character and pattern recognition, etc., can solve the problem of large quantization error, achieve the effect of small quantization error and reduce coding cost

Active Publication Date: 2017-09-15
NORTHWESTERN POLYTECHNICAL UNIV
View PDF3 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] In order to overcome the shortcomings of the existing static image lossy compression methods with large quantization errors, the

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
  • Dictionary learning static image lossy compression method based on minimum quantization error criterion
  • Dictionary learning static image lossy compression method based on minimum quantization error criterion
  • Dictionary learning static image lossy compression method based on minimum quantization error criterion

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0019] The specific steps of the dictionary learning static image lossy compression method based on the minimum quantization error criterion of the present invention are as follows:

[0020] 1. Image segmentation and normalization.

[0021] For all training images, they are first divided into 16×16 image blocks according to the raster scanning order, and the translation amount is 2 Then for each image block b i Carry out normalization with reference to formula (1)

[0022]

[0023] Finally, straighten the image patch into a vector Constitutes the input signal for dictionary learning.

[0024] 2. Image block clustering.

[0025] In order to ensure the purity of the learned dictionary, it is necessary to cluster the image blocks first, so that each class has similar image blocks. Since the training image is overlapped and divided into blocks, the number of image blocks obtained is very large, so firstly, 10% of the image blocks are randomly selected from all image bloc...

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 dictionary learning static image lossy compression method based on minimum quantization error criterion, and is to solve the problem of large quantification error of an existing static image lossy compression method. The technical scheme is characterized by adding information entropy of an index corresponding to each sparse coefficient into a sparse coding objective function as a regularization term; when selecting dictionary atoms through an orthogonal matching pursuit algorithm, minimizing the information entropy to limit dispersity of the dictionary atoms, thereby reducing coding cost of the index corresponding to each sparse coefficient; and meanwhile, in the dictionary learning process, ranking the sparse coefficients and finding k habit sequence division capable of enabling the total dispersion square sum of the sparse coefficients to be minimum, each division serving as one quantification group, different quantification groups adopting different quantization steps, the same quantification group adopting the same quantization steps, thereby enabling the final quantization error to be minimum.

Description

technical field [0001] The invention relates to a static image lossy compression method, in particular to a dictionary learning static image lossy compression method based on the minimum quantization error criterion. Background technique [0002] The document "Compressibility constrained sparse representation with learned dictionary for low bit-rate image compression, IEEE Transactions on Circuits and Systems for Video Technology, 2014, Vol24(10), p1743–1757" discloses a sparse coding method based on convex relaxation and compression constraints Lossy compression for images. This method replaces the traditional tracking matching algorithm with convex relaxation-based sparse coding, which strengthens the sparsity and stability of image representation coefficients. At the same time, the compression constraint is added to the solution process of sparse coding, and the sparse coding problem is transformed into The norm optimization problem approaches the optimal solution of t...

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): G06T9/00G06K9/62H03M7/30
CPCH03M7/3059H03M7/6041H03M7/6088G06T9/00G06F18/23213
Inventor 夏勇王昊张艳宁
Owner NORTHWESTERN POLYTECHNICAL UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products