CN105007490B - Jpeg compression algorithms based on OmapL138 chips - Google Patents
Jpeg compression algorithms based on OmapL138 chips Download PDFInfo
- Publication number
- CN105007490B CN105007490B CN201510479206.1A CN201510479206A CN105007490B CN 105007490 B CN105007490 B CN 105007490B CN 201510479206 A CN201510479206 A CN 201510479206A CN 105007490 B CN105007490 B CN 105007490B
- Authority
- CN
- China
- Prior art keywords
- omapl138
- chips
- data
- memories
- jpeg compression
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Compression Or Coding Systems Of Tv Signals (AREA)
- Compression Of Band Width Or Redundancy In Fax (AREA)
Abstract
The invention discloses a kind of Jpeg compression algorithms based on OmapL138 chips, comprise the following steps：（1）Image YUV420 data are read by the ARM kernels of OmapL138 chips, and are stored data into DDR2 memories；（2）ARM kernels read view data in batches from DDR2, and view data are linearly arranged in the L2 memories of the DSP core of OmapL138 chips；（3）DSP extracts view data from L2 memories, carries out dct transform and transformation results are quantified, and the result for calculating completion is sent back in L2 memories；（4）ARM from L2 memories after extraction process as a result, carry out entropy coding；（5）Repeat step（2）~（4）, the processing until completing all image datas, Jpeg compression completions.By the above-mentioned means, can be maximally utilized the performance of OmapL138 chips, allow Jpeg compression algorithms to realize that slow-footed problem is greatly improved, the practical value of Jpeg compression algorithms is greatly improved.
Description
Technical field
The present invention relates to a kind of Jpeg compression algorithms, are to be related to a kind of Jpeg based on OmapL138 chips specifically
Compression algorithm.
Background technology
Standard image file format in Windows operating system is BMP, if the points handled in real time are relatively more, place
Data volume after reason is still that very big, so big data volume may can not meet real-time Transmission, it is desirable that will figure
As data can carry out certain compression before transmission, on the premise of picture quality is not lost, data volume is reduced, more in time
Data are transmitted.
JPEG（Joint Photographic Expert Group, i.e. joint photographic experts group）In the end of the year 1986 into
It is vertical, it is that the compression algorithm of the static image of continuous tone formulates the world of encoding-decoding process and compressed image data format
Standard.The fundamental system of Joint Photographic Experts Group employs the method based on discrete cosine transform and Variable Length Code compression, is not influencing
On the premise of the subjective quality of image, using the teaching of the invention it is possible to provide larger compression ratio.Due to being performed the encoding operation again after discrete cosine transform
Distortion can be caused, corresponding distortion level and set compression ratio are closely related, therefore the image rebuild can not be very complete
Beauteously repeat original pictures, but the vision of people can not be distinguished well.
Joint Photographic Experts Group is easily realized in practical applications, particularly in terms of Real Time Compression, can be good at reaching real-time pressure
The requirement of contracting.
As shown in Figure 1, the DCT in face of whole image data in order to after the completion of is operated, it is necessary to first by original picture number
According to the unitization block of pixels for being divided into 8x8, then using scan mode from top to bottom, from left to right to the pixel after segmentation
Block rearranges order, after completing aforesaid operations, carries out the operation of dct transform, the official then provided by Joint Photographic Experts Group
Quantify table and quantization operation is carried out to 64 coefficients.
After DCT coefficient is quantified, the coefficient after quantization is re-started into sequence according still further to " Z " word sortord, uses stroke
Obtained bit stream is generated intermediate symbols sequence by coding, is encoded intermediate symbols using the Huffman methods encoded,
It can obtain corresponding compressed image code stream.Each block of pixels just completes whole JPEG pressures after carrying out above-mentioned processing in order
Compression process, after then doing corresponding flag code filling by the picture format of Joint Photographic Experts Group, you can file suffixes is after being compressed
.jpg image.
The main purpose of Jpeg compression algorithms is the storage and transmission for the ease of image, and pixel data is compressed,
Ensure the quality of picture at the same time.But since image data is complicated, operand is big, and when being compressed, speed is very slow, especially
When being handled in face of large image and high-volume picture, existing Jpeg compression algorithms are difficult to meet actual demand in speed,
There is an urgent need to a kind of high speed, real-time compression method.
The content of the invention
It is an object of the invention to provide a kind of Jpeg compression algorithms based on OmapL138 chips, solve in the prior art
The problem of Jpeg algorithms can not meet current social actual demand because caused by the big processing speed of data volume is slow.
To achieve these goals, the technical solution adopted by the present invention is as follows：
Jpeg compression algorithms based on OmapL138 chips, comprise the following steps：
（1）Image YUV420 data are read by the ARM kernels of OmapL138 chips, and store data into DDR2 memories
In；
（2）ARM kernels read view data in batches from DDR2, and view data is linearly arranged in OmapL138 chips
DSP core L2 memories in；
（3）DSP extracts view data from L2 memories, carries out dct transform and transformation results are quantified, and will calculate
The result of completion is sent back in L2 memories；
（4）ARM from L2 memories after extraction process as a result, carry out entropy coding；
（5）Repeat step（2）~（4）, the processing until completing all image datas, Jpeg compression completions.
Further, the step（1）In, ARM kernels are first filled the mark of Jpeg format specifications, insert figure
Wide, the high information of picture；Then the Huffman codings for needing to be frequently performed are fabricated to a Huffman table by ARM kernels, will
Zig-Zag arrangement modes are fabricated to a Zig-Zag concordance list.
Preferably, the step（2）In, the data volume arrangements such as the linear arrangement of view data progress uses.
Yet further, the step（3）Specific method it is as follows：
The data of 17696 Bytes are disposably extracted from the L2 memories of DSP by IDMA technologies into L1 memories, by
DSP handles the data in L1, finally returns the data transfer for handling completion in L2.
Preferably, data transfer mode is serial transmission mode in the DSP, and data processing method is serial process side
Formula.
Preferably, at least two buffering areas are provided with the L2 memories of the DSP, data transfer and data between L2 and L1
Processing is parallel.
Preferably, the step（3）In, DSP carries out dct transform using pure compilation mode.
The heretofore described OMAPL138 used is the C6748 Floating-point DSPs kernel and ARM9 kernels that TI companies release
Double-core high-speed processor, the device collection image, voice, network, be stored in one, cost-effective.In the chips, frequency highest
C6748 kernels up to 456MHz provide the fixed point ability to work of floating-point ability to work and higher performance, and ARM9 has height
Flexibility, developer can be convenient for its application addition man-machine interface, network on it using the operating system such as Linux
Function, touch-screen etc..OMAPL138 total power consumptions under different service conditions are 440mW, and standby mode power consumption is 15mW, its
Memory and peripheral resources very abundant, can meet the design requirement of Jpeg compressibilities completely, and can conveniently in the future into
The extension and upgrading of row system.
Compared with prior art, the invention has the advantages that：
（1）The present invention makes full use of the different qualities of OMAPL138 chip double-cores, by the use of ARM9 kernels as controller, carries
The view data of YUV is taken out, carries out logical process, and utilizes DSP to complete specific Image Data Compression computing, so that
Two core performances of OMAPL138 chips are not fully exerted, and then improve Jpeg compression algorithms and realize efficiency.
（2）The present invention carries out dct transform by the way of pure compilation, can effectively accelerate the execution speed of DSP, and not
Influence compression quality.
（3）The present invention sets multiple buffering area in dsp, and serial mould is used in simple data transfer and data processing
Formula carries out, and data processing and data transfer are carried out at the same time, and carries out parallel schema, both mutually take into account, and drastically increase
Compression efficiency, shortens compression time.
（4）The present invention fully merges the characteristic of OMAPL138 chips with the structure that Jpeg compresses with flow characteristic, mutually
Utilize, the maximization for realizing OMAPL138 chip performances embodies, and also so that the efficiency of Jpeg compressions greatly promotes, effectively carries
High accommodation and practical value.
Brief description of the drawings
Fig. 1 is the basic procedure of JPEG compression algorithm in the prior art.
Fig. 2 is the structure diagram that view data is linearly arranged in the present invention.
Fig. 3 is the schematic diagram of dct transform in the present invention.
Fig. 4 is the data dispatch schematic diagram between the L1 memories and L2 memories of DSP core in the present invention.
Embodiment
The invention will be further described with reference to the accompanying drawings and examples, and embodiments of the present invention include but not limited to
The following example.
Embodiment
As shown in Figure 2 to 4, the Jpeg compression algorithms disclosed by the invention based on OmapL138 chips, including following step
Suddenly：
（1）Image YUV420 data are read by the ARM kernels of OmapL138 chips, and store data into DDR2 memories
In；
（2）ARM kernels read view data in batches from DDR2, and view data is linearly arranged in OmapL138 chips
DSP core L2 memories in；
（3）DSP extracts view data from L2 memories, carries out dct transform and transformation results are quantified, and will calculate
The result of completion is sent back in L2 memories；
（4）ARM from L2 memories after extraction process as a result, carry out entropy coding；
（5）Repeat step（2）~（4）, the processing until completing all image datas, Jpeg compression completions.
Specifically, step is being performed（1）When, ARM is filled firstly for the mark of the format specification of Jpeg, and is filled out
Enter wide, the high information of image, the characteristics of Huffman codings and Zig-Zag arrangement modes that ARM is frequently performed for needs
The form of a Huffman table and Zig-Zag concordance lists is made, then data are all stored in DDR2 memories, to accelerate
Subsequent execution efficiency.
Afterwards, ARM is according to the required calculating data of DSP, extracts data in batches from DDR2, and according to linear structure into
Row arrangement, is stored in L2.All segment datas carry out equivalent arrangements, in order to be read in follow-up data handling procedure
Write.The size of segment data amount is determined by L1 memory sizes and actual conditions, in the present embodiment, is per segment data size
17696 Bytes。
Since the calculating speed of L1 in DSP is most fast, when carrying out dct transform, by IDMA technologies from L2 once
The data of 17696 Bytes are read into L1, DSP only needs to handle the data in L1, by data transfer after having handled
Return in L2 and store, then read the data of 17696 Bytes again into L1, reprocess, then pass handling result back L2 storages, such as
This circulation, untill all data are all disposed.When the data processing in L2 finishes and then is extracted from DDR2 another
Batch of data, continues above-mentioned data transfer and processing procedure, until the view data for needing to compress is compressed completely.
In data transfer and data handling procedure in dsp, the data transfer of L2 to L1 is serial mode, the number in L1
Also it is serial mode according to processing, and the data transfer of L2 to L1 and the data processing of L1 are carried out at the same time, and are parallel schema.Specifically
Dct transform is as shown in Figure 3 and Figure 4.
By the above-mentioned means, can be maximally utilized the performance of OmapL138 chips, make Jpeg compression algorithms real
Existing slow-footed problem is greatly improved, and the practical value of Jpeg compression algorithms is greatly improved.
Above-described embodiment is merely a preferred embodiment of the present invention, not limiting the scope of the invention, as long as using
The design principle of the present invention, and the change for carrying out non-creative work on this basis and making, should all belong to the present invention's
Within protection domain.
Claims (7)
1. the Jpeg compression algorithms based on OmapL138 chips, it is characterised in that comprise the following steps：
（1）Image YUV420 data are read by the ARM kernels of OmapL138 chips, and are stored data into DDR2 memories；
（2）ARM kernels read view data in batches from DDR2, and view data is linearly arranged in OmapL138 chips
In the L2 memories of DSP core；
（3）DSP extracts view data from L2 memories, carries out dct transform and transformation results are quantified, and completed calculating
Result send back in L2 memories；
（4）ARM from L2 memories after extraction process as a result, carry out entropy coding；
（5）Repeat step（2）~（4）, the processing until completing all image datas, Jpeg compression completions.
2. the Jpeg compression algorithms according to claim 1 based on OmapL138 chips, it is characterised in that the step
（1）In, ARM kernels are first filled the mark of Jpeg format specifications, insert wide, the high information of image；Then by ARM
The Huffman codings for needing to be frequently performed are fabricated to a Huffman table by core, and Zig-Zag arrangement modes are fabricated to one
Zig-Zag concordance lists.
3. the Jpeg compression algorithms according to claim 1 based on OmapL138 chips, it is characterised in that the step
（2）In, the data volume arrangements such as the linear arrangement of view data progress uses.
4. the Jpeg compression algorithms according to claim 1 based on OmapL138 chips, it is characterised in that the step
（3）Specific method it is as follows：
By IDMA technologies, disposably the data of 17696 Bytes of extraction are into L1 memories from the L2 memories of DSP, by DSP pairs
Data in L1 are handled, and finally return the data transfer for handling completion in L2.
5. the Jpeg compression algorithms according to claim 4 based on OmapL138 chips, it is characterised in that in the DSP
Data transfer mode is serial transmission mode, and data processing method is serial processing mode.
6. the Jpeg compression algorithms according to claim 5 based on OmapL138 chips, it is characterised in that the DSP's
At least two buffering areas are provided with L2 memories, data transfer and data processing are parallel between L2 and L1.
7. the Jpeg compression algorithms according to claim 6 based on OmapL138 chips, it is characterised in that the step
（3）In, DSP carries out dct transform using pure compilation mode.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510479206.1A CN105007490B (en) | 2015-08-03 | 2015-08-03 | Jpeg compression algorithms based on OmapL138 chips |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510479206.1A CN105007490B (en) | 2015-08-03 | 2015-08-03 | Jpeg compression algorithms based on OmapL138 chips |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105007490A CN105007490A (en) | 2015-10-28 |
| CN105007490B true CN105007490B (en) | 2018-05-15 |
Family
ID=54379980
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510479206.1A Active CN105007490B (en) | 2015-08-03 | 2015-08-03 | Jpeg compression algorithms based on OmapL138 chips |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105007490B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111159784B (en) * | 2019-12-17 | 2022-02-01 | 飞天诚信科技股份有限公司 | Method and device for preventing differential power analysis attack in chip |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101137062A (en) * | 2007-09-20 | 2008-03-05 | 四川长虹电器股份有限公司 | Video coding system dual-core cooperation encoding method with dual-core processor |
| CN201550234U (en) * | 2009-11-13 | 2010-08-11 | 河北工业大学 | ARM and DSP-based image processing device |
| CN102445267A (en) * | 2011-11-21 | 2012-05-09 | 嘉兴中科声学科技有限公司 | Test method for digitalized acoustics detection system |
| CN102882274A (en) * | 2012-09-07 | 2013-01-16 | 中国电力科学研究院 | Terminal fusion system applied to distribution network |
| CN103634565A (en) * | 2013-09-24 | 2014-03-12 | 北京环境特性研究所 | OMAP-based ultraviolet and visible light dual-channel image acquisition, processing and display system |
| CN104767957A (en) * | 2015-04-09 | 2015-07-08 | 南阳师范学院 | Video capture coding method, system and device based on embedded type dual-core processor |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2003174650A (en) * | 2001-09-27 | 2003-06-20 | Canon Inc | Image coding apparatus, variable-length coding apparatus, control apparatus, and method thereof |
-
2015
- 2015-08-03 CN CN201510479206.1A patent/CN105007490B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101137062A (en) * | 2007-09-20 | 2008-03-05 | 四川长虹电器股份有限公司 | Video coding system dual-core cooperation encoding method with dual-core processor |
| CN201550234U (en) * | 2009-11-13 | 2010-08-11 | 河北工业大学 | ARM and DSP-based image processing device |
| CN102445267A (en) * | 2011-11-21 | 2012-05-09 | 嘉兴中科声学科技有限公司 | Test method for digitalized acoustics detection system |
| CN102882274A (en) * | 2012-09-07 | 2013-01-16 | 中国电力科学研究院 | Terminal fusion system applied to distribution network |
| CN103634565A (en) * | 2013-09-24 | 2014-03-12 | 北京环境特性研究所 | OMAP-based ultraviolet and visible light dual-channel image acquisition, processing and display system |
| CN104767957A (en) * | 2015-04-09 | 2015-07-08 | 南阳师范学院 | Video capture coding method, system and device based on embedded type dual-core processor |
Non-Patent Citations (2)
| Title |
|---|
| 《基于Omap5910嵌入式平台的并行JPEG压缩算法》;程志，贾智平，郝凤琦;《计算机工程与设计》;20081030;第29卷(第16期);全文 * |
| OMAPL138的双核通信设计;林淦，刘建群，许东伟，李嘉健;《机床与液压》;20150105(第22期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105007490A (en) | 2015-10-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2021208247A1 (en) | Mimic compression method and apparatus for video image, and storage medium and terminal | |
| CN109842803A (en) | A kind of method and device of compression of images | |
| CN101917622B (en) | 14bit wide image compression hardware coder | |
| CN107154062A (en) | A kind of implementation method of WebP Lossy Compression Algorithms, apparatus and system | |
| CN104012093B (en) | Method for handling image | |
| CN105847800B (en) | Method for compressing image and system based on all phase discrete sine biorthogonal conversion | |
| CN104683801A (en) | Image compression method and device | |
| CN101431691B (en) | Fast parallel compression method for high dynamic range image | |
| Duan et al. | Optimizing JPEG quantization table for low bit rate mobile visual search | |
| CN103179393A (en) | Dram compression scheme to reduce power consumption in motion compensation and display refresh | |
| CN101867809A (en) | High-speed image compression VLSI (very Large Scale integration) encoding method and encoder based on pulse array | |
| CN101272489A (en) | Encoding and decoding device and method for video image quality enhancement | |
| CN108769694B (en) | Method and device for Alpha channel coding based on FPGA | |
| CN103246499A (en) | Device and method for parallelly processing images | |
| CN101511025B (en) | Method and apparatus for compressing and decompressing image | |
| CN102724499A (en) | Variable-compression ratio image compression system and method based on FPGA | |
| CN101415117A (en) | Transmission method for high presence image based on GPGPU | |
| CN110324668A (en) | The inverse transform method and device in transform method, decoding in image block coding | |
| CN105007490B (en) | Jpeg compression algorithms based on OmapL138 chips | |
| CN101626504A (en) | Method for high speed JPEG decoding | |
| US20240121439A1 (en) | Point cloud attribute information encoding method and apparatus, point cloud attribute information decoding method and apparatus, and related device | |
| CN104980764A (en) | Parallel coding/decoding method, device and system based on complexity balance | |
| CN117560511A (en) | Spacer image compression method and system based on graph segmentation technology and electric power inspection | |
| CN111815502B (en) | FPGA acceleration method for multi-graph processing based on WebP compression algorithm | |
| CN108900842A (en) | Y data compression processing method, device, equipment and WebP compressibility |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |