CN105227953A - Image coding device, image-decoding device and its correlative coding and coding/decoding method - Google Patents
Image coding device, image-decoding device and its correlative coding and coding/decoding method Download PDFInfo
- Publication number
- CN105227953A CN105227953A CN201410257958.9A CN201410257958A CN105227953A CN 105227953 A CN105227953 A CN 105227953A CN 201410257958 A CN201410257958 A CN 201410257958A CN 105227953 A CN105227953 A CN 105227953A
- Authority
- CN
- China
- Prior art keywords
- image
- coding
- expansion mode
- header
- identifying information
- 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.)
- Pending
Links
Abstract
The invention provides a kind of image coding device, wherein comprise an expansion module, a quantization modules and a coding module.A target expansion mode, in order to for an image to be encoded, is selected in multiple expansion mode by this expansion module system, and according to this target expansion mode, one coefficient matrix is expanded into a quantification form.This quantization modules imposes a quantification program, to produce a quantized result according to the image data matrix of this quantification form to this image to be encoded.This quantized result is encoded to produce a coding result by this coding module, and will point out that an identifying information of this target expansion mode writes in this coding result.
Description
Technical field
The present invention is relevant to image coding/decoding technique, and especially relevant to the quantification technique in image coding/decoding program.
Background technology
Along with the progress of mechanics of communication, digital television broadcasting is gradually ripe, universal.Except transmitting via cable line, digital television signal also can pass through the equipment such as base station or artificial satellite and is passed with the kenel of wireless signal.In order to take into account improving picture quality and reduce the demand of transmitted data amount, transmission end usually can by image to be passed and sound signal encoding, compression.Accordingly, receiving terminal must correctly by receive signal decoding, decompression, the beginning can reduce video-audio signal.
Fig. 1 (A) presents the local function calcspar of an image coding system.In frame, prediction (intra-prediction) module 12 carries out predictor in frame for each image block in a video signal frame, respectively to produce its luma residual value (residual) matrix.The luma residual value matrix that in frame, prediction module 12 is selected is provided to discrete cosine transform (discretecosinetransform, DCT) module 14, carries out DCT program, to produce a DCT coefficient matrix.In order to reduce data volume further, two times transfer (secondarytransform) module 16 can be responsible for imposing two times transfer to the low-frequency component in this DCT coefficient matrix.Subsequently, the low-frequency component after two times transfer and other high frequency DCT coefficients without two times transfer can be recombined in quantization modules 18, and impose quantification program.When carrying out quantification program, required quantification form (quantizationtable) is be stored in memory 15.Quantizing form is a quantization weight value matrix, and its size is identical with DCT coefficient matrix.Easy speech it, if the size of DCT coefficient matrix that DCT module 14 exports is NxN, the size quantizing form also can be NxN.
In order to save storage space, some image coding system quantizing form size comparatively large (such as 32x32) can change the framework adopted as shown in Fig. 1 (B).In this framework, be stored in memory 15 and incomplete quantification form, but the coefficient matrix that a size is less.When quantization modules 18 needs to quantize form, launch module 17 and to fill up the mode such as (flatpadding) or bilinear interpolation (bilinearinterpolation), this small size coefficient matrix can be expanded into large-sized quantification form.In current art, launch module 17 and be designed to adopt a kind of fixing expansion mode (such as fixing use bilinear interpolation).The shortcoming of this way is that the variability of the quantification form produced accordingly is quite limited.
Summary of the invention
For solving the problem, the present invention proposes a kind of new image coding device and image-decoding device.Launch module by order and multiple different quantification form expansion mode is provided, the change elasticity quantizing table content can be increased according to image coding device of the present invention and image-decoding device.Larger-size image coding/the decode system of various quantification form can be applied to according to image coding device of the present invention and image-decoding device, such as digital audio/video encoding and decoding technique standard (audiovideocodingstandard, AVS) system.
A specific embodiment according to the present invention is a kind of image coding device, wherein comprises an expansion module, a quantization modules and a coding module.A target expansion mode, in order to for an image to be encoded, is selected in multiple expansion mode by this expansion module system, and according to this target expansion mode, one coefficient matrix is expanded into a quantification form.This quantization modules system is in order to impose a quantification program, to produce a quantized result according to the image data matrix of this quantification form to this image to be encoded.This coding module, in order to encode this quantized result to produce a coding result, and will point out that an identifying information of this target expansion mode writes in this coding result.
Another specific embodiment according to the present invention is a kind of image-decoding device, wherein comprises a decoder module, and launches module and an inverse quantization module.This decoder module system, in order to the coding result decoding by an image to be decoded, to produce a decoded result, and reads out an identifying information in this decoded result.One coefficient matrix in order to select a target expansion mode according to this identifying information in multiple expansion mode, and is expanded into a quantification form according to this target expansion mode by this expansion module system.This inverse quantization module system imposes a re-quantization program in order to quantize rear image data matrix according to this quantification form to one of this image to be decoded.
Another specific embodiment according to the present invention is a kind of video encoding method.First, for an image to be encoded, a target expansion mode is selected in multiple expansion mode.According to this target expansion mode, it is a quantification form that a coefficient matrix is unfolded.According to this quantification form, an image data matrix of this image to be encoded is subjected to a quantification program, to produce a quantized result.This quantized result is encoded to produce a coding result.Point out that an identifying information of this target expansion mode is written in this coding result.
Another specific embodiment according to the present invention is a kind of image decoding method.First, a coding result of an image to be decoded is decoded, to produce a decoded result.One identifying information is read out in this decoded result.According to this identifying information, a target expansion mode is selected in multiple expansion mode.According to this target expansion mode, it is a quantification form that a coefficient matrix is unfolded.According to this quantification form, after one of this image to be decoded quantizes, image data matrix is subjected to a re-quantization program.
Can be further understood by following detailed Description Of The Invention and institute's accompanying drawings about the advantages and spirit of the present invention.
Accompanying drawing explanation
Fig. 1 (A) and Fig. 1 (B) presents the local function calcspar of an image coding system.
Fig. 2 is the functional block diagram according to the image coding system in one embodiment of the invention.
Fig. 3 is the functional block diagram according to the image-decoding system in one embodiment of the invention.
Fig. 4 is the flow chart according to the video encoding method in one embodiment of the invention.
Fig. 5 is the flow chart according to the image decoding method in one embodiment of the invention.
Should be noted that, of the present inventionly graphicly comprise the functional block diagram presenting multiple functional module associated with each other.These graphic not thin portion circuit diagrams, and connecting line is wherein only in order to represent signal stream.Multiple interactive relationship between functional assembly and/or program not necessarily will begin to reach through direct electrically connect.In addition, the function of individual elements not necessarily will the mode as illustrated in graphic be distributed, and distributed block not necessarily will realize by electronic building brick in a distributed manner.
Embodiment
A specific embodiment according to the present invention is a kind of image coding system, and its functional block diagram system is illustrated in Fig. 2.Image coding system 200 comprises prediction module 22 in frame, discrete cosine transform module 24, memory 25, two times transfer module 26, launches module 27, quantization modules 28 and coding module 29.In practical application, image coding system 200 can individualism, also can be incorporated into the larger-size image coding/decode system of various quantification form, such as but not limited to digital audio/video encoding and decoding technique standard (AVS) system.
In frame, prediction module 22 carries out predictor in frame for each image block in a video signal frame respectively, to produce its luma residual value matrix.Then, the luma residual value matrix that in frame, prediction module 22 exports is provided to discrete cosine transform (DCT) module 24 and carries out DCT program, to produce a DCT coefficient matrix.Two times transfer module 26 is responsible for imposing two times transfer to the low-frequency component in DCT coefficient matrix.Subsequently, the low-frequency component after two times transfer and other high frequency DCT coefficients without two times transfer are recombined in quantization modules 28, and are subjected to quantification program according to a quantification form.A referential matrix number is stored in memory 25.Launch module 27 to be responsible for this referential matrix number being expanded into the relatively large quantification form of size, for quantization modules 28.
Be different from prior art, launch the image to be encoded that module 27 can process at present for image coding system 200, in multiple expansion mode, select a target expansion mode, and according to this target expansion mode, this referential matrix number is expanded into quantification form.For example, this multiple expansion mode can comprise and homogeneously fills up (flatpadding) formula deployment schemes, bilinear interpolation (bilinearinterpolation) deployment schemes, hybrid deployment schemes or the self-defined deployment schemes of a user.So-called hybrid deployment schemes means launches module 27 certain a part of one first expansion mode (such as homogeneous fill up) for referential matrix number, and for another part one second expansion mode (such as bilinear interpolation) of referential matrix number.So-called User Defined deployment schemes means that a set of matrix of circuit designers designed, designed launches rule, and its rule also can be stored in memory 25.In practice, launch module 27 select expansion mode according to can including but not limited to whether wishing to retain more high frequency detail for this image to be encoded.Launch module by order and provide multiple different quantification form expansion mode, image coding device 200 can have good quantification form content change elasticity.
Should be noted that, the producing method of the coefficient matrix in memory 25, and the detailed embodiment of various deployment schemes is known to persond having ordinary knowledge in the technical field of the present invention, repeats no more in this.In addition, category of the present invention is not defined in specific storage mechanism.Memory 25 can comprise one or more volatility or nonvolatile memory device, such as random-access semiconductor memory, read-only memory, magnetic and/or optical memory, flash memory etc.
Subsequently, the quantized result that quantization modules 28 produces is provided to coding module 29.This quantized result is encoded to produce a coding result by coding module 29, and will point out that an identifying information of this target expansion mode writes in this coding result.This identifying information (such as give each and quantize form expansion mode one particular number) is be supplied to coding module 29 by expansion module 27.In an embodiment, coding module 29 is by the header (header) of this identifying information write coding result.For example, this header can be an image sequence header, an image header or a fragment image (slice) header.For the situation by expansion mode identifying information write image header, it is all adopt same expansion mode that this identifying information can represent whole the image corresponding to this header.In an embodiment, this identifying information can be designed to point out that same quantification form (being namely applicable to same expansion mode) shared by multiple image to be encoded further.For example, this identifying information can be pointed out to be started at by this image to be encoded current, and other multiple images to be encoded of process are also applicable to same quantification form expansion mode by image coding system 200 subsequently.In that case, after, the coding result of multiple image to be encoded does not just need to comprise expansion mode identifying information.
Should be noted that, the input signal that category of the present invention is not defined in quantization modules 28 is necessary for DCT coefficient matrix or its two times transfer result, but contains various image data matrix.More particularly, concept of the present invention can be applicable to the various image coding system needing to be launched by small size referential matrix number to become large scale quantification form.In addition, persond having ordinary knowledge in the technical field of the present invention can understand, and separately has multiple circuit configurations and assembly can realize concept of the present invention without departing from the spirit of the invention.
Another specific embodiment according to the present invention is a kind of image-decoding device, and its functional block diagram system is illustrated in Fig. 3.Image-decoding device 300 comprises decoder module 31, inverse quantization module 32, and launches module 33, memory 34, against two times transfer module 35, inverse discrete cosine modular converter 36, and one rebuilds module 37.Decoder module 31 is in order to the coding result decoding by an image to be decoded, to produce a decoded result, and reads out an identifying information in this decoded result.A referential matrix number is stored in memory 34.This identifying information is provided to and launches module 33.Launch module 33 to be responsible in multiple expansion mode, selecting a target expansion mode according to this identifying information, and according to this target expansion mode, this referential matrix number is expanded into a quantification form.In practice, the storage location of identifying information in decoded result is decoder module 31 known (being ordered by coding side and decoding end agreement), identifying information and corresponding the closing also in advance for launching known to module 33 quantizing form expansion mode in advance.For example, this multiple expansion mode can comprise homogeneous filled type deployment schemes, bilinear interpolation deployment schemes, hybrid deployment schemes or the self-defined deployment schemes of a user.The rule of User Defined deployment schemes can be stored in memory 34.If launch module 33 to judge that identifying information points out that current image system to be decoded corresponds to User Defined deployment schemes, launch module 33 just can capture the self-defined deployment schemes of this user operation regulation from memory 34.
Quantize form according to after the expansion that expansion module 33 provides, after inverse quantization module 32 quantizes one of this image to be decoded, image data matrix imposes a re-quantization program, to produce a de-quantizing results.Inverse two times transfer module 35 is in order to impose inverse two times transfer program to the low frequency component in this de-quantizing results, to produce a low frequency component submatrix.Subsequently, the high-frequency components in this low frequency component submatrix and this de-quantizing results is provided to inverse discrete cosine modular converter 36, carries out inverse discrete cosine conversion program, to produce a residual value matrix.Rebuilding module 37 is in order to according to this residual value matrix and many reference image datas, rebuilds this image block.
Similar to image coding system 200, the input signal that category of the present invention is not defined in inverse quantization module 32 is necessary for the DCT coefficient matrix after quantification or its two times transfer result, but contains the image data matrix after various quantification.More particularly, concept of the present invention can be applicable to the various image-decoding system needing to be launched by small size referential matrix number to enclose large scale quantification form.
Another specific embodiment according to the present invention is a kind of video encoding method, and its flowchart is illustrated in Fig. 4.First, step S41 is for an image to be encoded, selects a target expansion mode in multiple expansion mode.Step S42 is according to this target expansion mode, a coefficient matrix is expanded into a quantification form.Then, step S43 is according to this quantification form, imposes a quantification program, to produce a quantized result to an image data matrix of this image to be encoded.In step S44, this quantized result is encoded to produce a coding result, and points out that an identifying information of this target expansion mode is written in this coding result.Persond having ordinary knowledge in the technical field of the present invention can understand, and the various operations previously described when introducing image coding device 200 change the video encoding method that also can be applied in Fig. 4, and its details repeats no more.
Another specific embodiment according to the present invention is a kind of image decoding method, and its flowchart is illustrated in Fig. 5.First, step S51 is by a coding result decoding of an image to be decoded, to produce a decoded result.Step S52 for read out an identifying information in this decoded result.Step S53 is according to this identifying information, selects a target expansion mode in multiple expansion mode.Step S54 is according to this target expansion mode, a coefficient matrix is expanded into a quantification form.In step S55, after one of this image to be decoded quantizes, image data matrix is subjected to a re-quantization program according to this quantification form.Persond having ordinary knowledge in the technical field of the present invention can understand, and the various operations previously described when introducing image-decoding device 300 change the image decoding method that also can be applied in Fig. 5, and its details repeats no more.
By the above detailed description of preferred embodiments, be that hope clearly can describe feature of the present invention and spirit, and not with above-mentioned disclosed preferred embodiment, category of the present invention limited.On the contrary, its objective is wish to contain various change and tool equality be arranged in the present invention institute in the category of the scope of the claims applied for.
Claims (20)
1. an image coding device, comprises:
One launches module, in order to for an image to be encoded, selects a target expansion mode in multiple expansion mode, and according to this target expansion mode, one coefficient matrix is expanded into a quantification form;
One quantization modules, in order to impose a quantification program, to produce a quantized result according to the image data matrix of this quantification form to this image to be encoded; And
One coding module, in order to encode this quantized result to produce a coding result, and will point out that an identifying information of this target expansion mode writes in this coding result.
2. image coding device as claimed in claim 1, it is characterized in that, this identifying information writes in a header of this coding result by this coding module system.
3. image coding device as claimed in claim 2, it is characterized in that, this header is an image sequence header, an image header or a fragment image header.
4. image coding device as claimed in claim 1, it is characterized in that, this identifying information points out that this quantification form shared by multiple image to be encoded further.
5. image coding device as claimed in claim 1, it is characterized in that, this multiple expansion mode comprises a homogeneous filled type deployment schemes, a bilinear interpolation deployment schemes, a hybrid deployment schemes or the self-defined deployment schemes of a user.
6. an image-decoding device, comprises:
One decoder module, in order to the coding result decoding by an image to be decoded, to produce a decoded result, and reads out an identifying information in this decoded result;
One launches module, in order to select a target expansion mode according to this identifying information in multiple expansion mode, and according to this target expansion mode, one coefficient matrix is expanded into a quantification form; And
One inverse quantization module, imposes a re-quantization program in order to quantize rear image data matrix according to this quantification form to one of this image to be decoded.
7. image-decoding device as claimed in claim 6, it is characterized in that, this decoder module system reads out this identifying information in a header of this coding result.
8. image-decoding device as claimed in claim 7, it is characterized in that, this header is an image sequence header, an image header or a fragment image header.
9. image-decoding device as claimed in claim 6, it is characterized in that, this identifying information points out that this quantification form shared by multiple image to be encoded further.
10. image-decoding device as claimed in claim 6, is characterized in that, this multiple expansion mode comprises a homogeneous filled type expansion, a bilinear interpolation expansion, a hybrid deployment schemes or the self-defined deployment schemes of a user.
11. 1 kinds of video encoding methods, comprise:
A (), for an image to be encoded, selects a target expansion mode in multiple expansion mode;
B one coefficient matrix is expanded into a quantification form according to this target expansion mode by ();
C (), according to this quantification form, imposes a quantification program, to produce a quantized result to an image data matrix of this image to be encoded; And
D this quantized result is encoded to produce a coding result by (), and will point out that an identifying information of this target expansion mode writes in this coding result.
12. video encoding methods as claimed in claim 11, it is characterized in that, in step (d), this identifying information is written in a header of this coding result.
13. video encoding methods as claimed in claim 12, is characterized in that, this header is an image sequence header, an image header or a fragment image header.
14. video encoding methods as claimed in claim 11, is characterized in that, this identifying information points out that this quantification form shared by multiple image to be encoded further.
15. video encoding methods as claimed in claim 11, is characterized in that, this multiple expansion mode comprises a homogeneous filled type deployment schemes, a bilinear interpolation deployment schemes, a hybrid deployment schemes or the self-defined deployment schemes of a user.
16. 1 kinds of image decoding methods, comprise:
A one coding result of one image to be decoded is decoded, to produce a decoded result by ();
B () reads out an identifying information in this decoded result;
C () selects a target expansion mode according to this identifying information in multiple expansion mode;
D one coefficient matrix, according to this target expansion mode, is expanded into a quantification form by (); And
E () quantizes rear image data matrix according to this quantification form to one of this image to be decoded and imposes a re-quantization program.
17. image decoding methods as claimed in claim 16, is characterized in that, in step (b), this identifying information system is read out in a header of this coding result.
18. image decoding methods as claimed in claim 17, is characterized in that, this header is an image sequence header, an image header or a fragment image header.
19. image decoding methods as claimed in claim 16, is characterized in that, this identifying information points out that this quantification form shared by multiple image to be encoded further.
20. image decoding methods as claimed in claim 16, is characterized in that, this multiple expansion mode comprises a homogeneous filled type expansion, a bilinear interpolation expansion, a hybrid deployment schemes or the self-defined deployment schemes of a user.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410257958.9A CN105227953A (en) | 2014-06-11 | 2014-06-11 | Image coding device, image-decoding device and its correlative coding and coding/decoding method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410257958.9A CN105227953A (en) | 2014-06-11 | 2014-06-11 | Image coding device, image-decoding device and its correlative coding and coding/decoding method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105227953A true CN105227953A (en) | 2016-01-06 |
Family
ID=54996594
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410257958.9A Pending CN105227953A (en) | 2014-06-11 | 2014-06-11 | Image coding device, image-decoding device and its correlative coding and coding/decoding method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105227953A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2006001533A2 (en) * | 2004-06-29 | 2006-01-05 | Olympus Corporation | Methods and apparatus for image encoding/decoding, and extended image compression expansion |
| CN101056402A (en) * | 2006-02-15 | 2007-10-17 | 宏正自动科技股份有限公司 | Image transmission system |
| CN103238328A (en) * | 2010-12-09 | 2013-08-07 | 索尼公司 | Image processing device and image processing method |
| CN103780912A (en) * | 2011-02-10 | 2014-05-07 | 索尼公司 | Image processing device and image processing method |
-
2014
- 2014-06-11 CN CN201410257958.9A patent/CN105227953A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2006001533A2 (en) * | 2004-06-29 | 2006-01-05 | Olympus Corporation | Methods and apparatus for image encoding/decoding, and extended image compression expansion |
| CN101056402A (en) * | 2006-02-15 | 2007-10-17 | 宏正自动科技股份有限公司 | Image transmission system |
| CN103238328A (en) * | 2010-12-09 | 2013-08-07 | 索尼公司 | Image processing device and image processing method |
| CN103780912A (en) * | 2011-02-10 | 2014-05-07 | 索尼公司 | Image processing device and image processing method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US8666179B2 (en) | Image encoding apparatus and decoding apparatus | |
| KR101377021B1 (en) | Encoding device and method, decoding device and method, and transmission system | |
| US9509988B2 (en) | Motion video encoding apparatus, motion video encoding method, motion video encoding computer program, motion video decoding apparatus, motion video decoding method, and motion video decoding computer program | |
| CN110300301B (en) | Image coding and decoding method and device | |
| US20170064335A1 (en) | Method and apparatus for image transformation, and method and apparatus for image inverse-transformation based on scanning sequence | |
| US20220188633A1 (en) | Low displacement rank based deep neural network compression | |
| CN113994348A (en) | Linear neural reconstruction for deep neural network compression | |
| JP2023543985A (en) | Template matching prediction for versatile video coding | |
| CN113330747A (en) | Method and apparatus for video encoding and decoding using bi-directional optical flow adaptive to weighted prediction | |
| CN112806011A (en) | Improved virtual time affine candidates | |
| JP2022534999A (en) | Conversion selection for implicit multiple conversion selection | |
| CN113728637A (en) | Framework for encoding and decoding low rank and shifted rank based layers of deep neural networks | |
| CN105227953A (en) | Image coding device, image-decoding device and its correlative coding and coding/decoding method | |
| CN117256142A (en) | Method and apparatus for encoding/decoding images and video using artificial neural network based tools | |
| CN116018805A (en) | KARHENEN LOEVE transform for video coding | |
| CN115516858A (en) | Zoom list control in video coding | |
| CN115280786A (en) | Advanced constraint flags for local chroma quantization parameter control | |
| US20140152767A1 (en) | Method and apparatus for processing video data | |
| CN114127746A (en) | Compression of convolutional neural networks | |
| TWI555388B (en) | Image encoding apparatus, image decoding apparatus and encoding and decoding methods thereof | |
| CN112806010A (en) | Method and apparatus for video encoding and decoding using predictor candidate list | |
| CN113473154B (en) | Video encoding method, video decoding method, video encoding device, video decoding device and storage medium | |
| JP2002223455A (en) | Image coding method and device, and image decoding method and device | |
| CN115362679A (en) | Method and apparatus for video encoding and decoding | |
| CN105323589A (en) | Coding and decoding methods and coding and decoding devices for video system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20160106 |
|
| WD01 | Invention patent application deemed withdrawn after publication |