High-performance optimization method for JPG format picture

An optimization method and high-performance technology, applied in the field of image processing, can solve the problems of long time consumption, affecting the number of concurrent, insufficient CPU, etc., to achieve the effect of efficient processing efficiency, reduction of optimization time, and reduction of memory allocation

Inactive Publication Date: 2019-11-05
杭州雄迈集成电路技术股份有限公司
View PDF1 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] 1. Image encoding and decoding is a very CPU-intensive job. Under massive concurrency, it is easy to cause insufficient CPU, affect the number of concurrency, and reduce business processing capabilities
[0006] 2. Image codec is a difficult job to develop, and many Internet companies lack the motivation for in-depth optimization
[0007] 3. The whole process of image encoding and decoding takes a long time, and the CPU load is heavy

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
  • High-performance optimization method for JPG format picture
  • High-performance optimization method for JPG format picture
  • High-performance optimization method for JPG format picture

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0041] A high-performance optimization method for JPG format pictures, the method includes the following steps:

[0042] Step 1): Read the original JPG format picture, take out the Huffman code table, quantization table, picture width and height, and the start address of the compressed macroblock. Generally, every 16x16 pixel on the JPG format picture is divided into a macroblock;

[0043] Step 2): Carry out entropy decoding on the macroblock to obtain the quantized DCT coefficients;

[0044] The entropy decoding process is:

[0045] a) First, the macroblock is divided into several data units DU according to luminance and chrominance, and the data in each data unit DU is divided into two types, one is DC data that has been differentially coded, and the other is run-length coded For AC data, DPCM is used for differential coding, and RLE is used for run length coding;

[0046] b) Perform a normal Huffman decoding on the two kinds of data first, and then perform DPCM and RLE decoding resp...

Embodiment 2

[0061] The comparison test table of the optimization result of the embodiment 2 and the optimization result of the traditional technology:

[0062] Optimization operation hours FFmpeg 85ms ImagicMagick 59ms Libjpeg 40ms this invention 7ms

[0063] It can be seen that the actual effect of the present invention reduces the CPU usage time by more than 80% compared with the traditional technology.

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 high-performance optimization method for a JPG format picture, and the method comprises the steps: an original JPG format picture is read, and a Huffman code table, a quantization table, the width and height of the picture, and a compressed macro block initial address are taken out; entropy decoding is carried out on the macro block, and a quantized DCT coefficient is solved; inverse quantization is carried out on the quantized DCT coefficient, and an actual DCT coefficient is restored; the quantization table with the highest compression ratio is selected, and the actual DCT coefficient is quantized again; entropy coding is carried out on the quantized DCT coefficient; according to the file format is JPG file format, new quantization and Huffman code tables are filled, a new JPG format picture is generated. According to the method, complex operation in a traditional method is removed, the performance is improved, memory allocation is reduced, a data structurecannot be destroyed when the pictures are continuously optimized, compared with industrial-grade libraries such as libjpeg in the prior art, the method has higher processing efficiency, and the optimization time of one picture is shortened by 80% compared with the traditional technology.

Description

Technical field [0001] The invention relates to a high-performance optimization method for pictures, more specifically, to a high-performance optimization method for JPG format pictures, which belongs to the field of picture processing. Background technique [0002] JPG is one of the most widely used image formats. Usually, the original JPG generated from digital cameras, mobile phones, etc. has the problem of being too large. In order to store and transmit, many applications need to optimize its size to reduce the file size. volume. Some professional websites have dedicated servers to optimize media data. Under the pressure of massive business, improving the performance of image compression and optimization has important practical significance. [0003] JPG optimization involves the underlying decoding and encoding technology, which is difficult to implement directly from the underlying. The existing optimization methods are basically implemented using the following three third-...

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): H04N19/176H04N19/91H04N19/124H04N19/625H04N19/182H04N19/186H04N19/48
CPCH04N19/124H04N19/176H04N19/182H04N19/186H04N19/48H04N19/625H04N19/91
Inventor 王仁毅肖文勇
Owner 杭州雄迈集成电路技术股份有限公司
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

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap