CN1342369A - Improved cascaded compression method and system for digital video and images - Google Patents
Improved cascaded compression method and system for digital video and images Download PDFInfo
- Publication number
- CN1342369A CN1342369A CN00804549A CN00804549A CN1342369A CN 1342369 A CN1342369 A CN 1342369A CN 00804549 A CN00804549 A CN 00804549A CN 00804549 A CN00804549 A CN 00804549A CN 1342369 A CN1342369 A CN 1342369A
- Authority
- CN
- China
- Prior art keywords
- cascade
- quantizer
- compressibility
- expectation
- quantization error
- 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.)
- Granted
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
- H04N19/102—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or selection affected or controlled by the adaptive coding
- H04N19/124—Quantisation
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/40—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using video transcoding, i.e. partial or full decoding of a coded input stream followed by re-encoding of the decoded output stream
Abstract
A system and method are disclosed for reducing quantization error in a system using a cascaded compression scheme. An expected quantization error introduced by the cascaded compression scheme is determined. The expected quantization errors based upon two or more second or higher stage test quantizers are then compared. Based on this comparison, a quantizer for the second or higher stage of compression is selected to reduce and/or minimize the quantization error caused by the cascaded compression scheme.
Description
Invention field
The invention belongs to the video compression technique area, specifically, invention relates to the quantization error that causes is compressed in reduction owing to the cascade of digital video and image system and method.
Background of invention
Video telephone, Digital Television, teleconference and information superhighway are a few examples that digital age emerges in large numbers.The development of digital picture and Video processing has promoted the arriving of digital age.Particularly the digital image compression method has played critical effect in this development.Image compression has reduced the required data volume of expression digital picture.For example, can compress, obtain the accurate expression of original picture then by decompression colour, gray scale or black and white image.
Compression is carried out in storage or before sending data usually.This makes bulk information store and/or to transmit rapidly with a kind of mode of economy.As is generally known image compression is a two-way process normally, relates to compression and decompression.These processes may not be symmetrical, and promptly a used time and/or the computing capability of process may be different from another process for the compression algorithm that adopts given type.
Two class image compression are arranged usually: that diminish and harmless.In lossy compression method, the image through decompressing is just similar rather than identical with original picture.This is because of having at least a part to change in the initial data or having abandoned.Lossy compression comprises double sampling, differential pulse code modulation (DPCM) and discrete cosine transform (DCT) coefficient quantization.On the contrary, lossless compress has kept all data of original picture, promptly is the cataloged procedure of a completely reversibility in essence.Lossless compressiong comprises variable length code (VLC) and run length coding, RLC (RLC).
Compression ratio is normally defined the ratio of the data that obtain after the data content that needs compression and the compression.The lossy compression method method can provide 100: 1 above compression ratios.Lossless compression method can only reach the compression ratio about 3: 1 usually.Usually, along with the increase of lossy compression method ratio, the image degradation degree also increases.
Can reach given compression ratio with single stage compress or multi-stage compression.Image or video source just have been subjected to the cascade compression when the source signal of input is subjected to the multi-stage compression of series system.For example, in the cascade compression, at first with certain compression ratio pressure source signal (image or video), and then make compressed data be subjected to one second or back level compression, so that reach a bigger compression ratio.In fact, for the transmission of high efficiency storage or limited bandwidth system, may need bigger compression.
Joint Photographic Experts Group (JPEG) and motion picture expert group (MPEG) standard are the schemes of respectively image and video being compressed that very extensively adopts.Joint Photographic Experts Group is used for the colour of real world true to nature or gray level image are compressed.Though Joint Photographic Experts Group comprises and diminishes pattern that it is generally used for diminishing pattern, so that reach bigger compression ratio harmless.Usually, an imagery exploitation discrete cosine transform (DCT) transforms to frequency domain.The less frequency component of resulting its intermediate value is rejected, and the bigger component of the value of staying only.The component that these values are bigger is encoded with differential pulse code modulation (DPCM) and Huffman code then.The controllability of JPEG compression makes can consider variable compression ratio and fine tuning algorithm for a specific application requirements.
Mpeg standard adopts DCT and huffman coding method in conjunction with inter-frame coding, can obtain bigger compression ratio.MPEG-1 and MPEG-2 are respectively applied for low definition image sequence and high definition sequence usually.MPEG-4 at be the sound-as unified rather than frame of object and scene.MPEG-7 helps sound-as the location of content.
Above-mentioned compress technique adopts dct transform the DCT coefficient to be quantized and variable length code again, to reach data compression.Quantization DCT coefficient makes these compress techniques become and diminishes.As mentioned above, the lossy compression method scheme is that the data after decompressing are not a kind of compression schemes of the strictness copy of initial data.For example, in JPEG or MPEG compression scheme, every grade of compression compressing in cascade in fact all diminishes.In addition, carry out multi-stage compression and also introduced additional loss.
For this additional loss of illustration, have a look two kinds of lossy compression method situations (a) shown in Figure 1 and (b).In situation (a), source data 10 is compressed with compression ratio by compressibility 11 at 20: 1.In situation (b), source data 10 at first is subjected to the compression of compressing the cascade compressibility 12 that is subjected to compression in partial 2: 1 again at 10: 1 of the first order.In situation (b), second compression stage does not insert original source data 10, and just obtains the compressed signal of first compression stage output.Situation (a) and (b) reached compression in same 20: 1.Yet because the cascade compression, the mean square error (MSE) that situation (b) is introduced will be always more than or equal to situation (a).This additional error partly is because second or the selection of quantized value in the compression stage of back.
Some traditional cascade compressibility adopts one group of quantizing factor that does not have degeneration for each compression stage in succession.Utilization such as the numerical relation of q value=K* (3**n) is selected this group quantizing factor.If use one of them quantizer, can use any other quantizer in the stage subsequently then in the phase I.Yet these systems do not provide any and how to select the quantizer in the subsequent stage so that the cascade quantization error is reduced to minimum experience.Therefore, in this technical field, exist the requirement of improvement system, method and technology, reduce to minimum, particularly loss is reduced to minimum by the quantizer of second in suitable selection cascade compression scheme or back level with the MSE that the cascade compression is introduced.
Brief summary of the invention
An object of the present invention is to disclose the shortcoming of traditional cascade compressibility discussed above.
Another object of the present invention provides the method that the error of introducing is compressed in a kind of cascade of calculating given a pair of quantizer.
Another purpose of the present invention provides a kind of loss of the cascade compression being introduced by suitable selection quantizer and reduces to minimum method.
A preferred embodiment relates to the quantization error of introducing in the framework that is reduced in JPEG and MPEG compression scheme.
One aspect of the present invention relates to a kind of method that is used for the cascade compressibility, and this method comprises the following steps: to determine second or the expectation quantization error introduced of back level by the cascade compressibility; And to be used for second or the expectation quantization error of at least two quantizers of back level compare.This method also comprises according to comparative result selects one of them quantizer, makes the expectation quantization error of cascade compressibility reduce to minimum.
In a preferred embodiment of the invention, determine the expectation quantization error with a probability-distribution function.
Another aspect of the present invention relates to a kind of medium and equipment of carrying out said method.
To be illustrated these and some other embodiment of the present invention and situation in the detailed description below.
Brief Description Of Drawings
With reference to understanding characteristic of the present invention and advantage well below in conjunction with the accompanying drawing detailed description of preferred embodiments.In these accompanying drawings:
Fig. 1 is the schematic diagram of a non-cascade compressibility and a cascade compressibility;
Fig. 2 is the schematic diagram that is illustrated in the quantification situation of two quantizers in the cascade compressibility;
Fig. 3 is the schematic diagram that is illustrated in a quantization error in the cascade compressibility;
Fig. 4 is the block diagram according to a typical computer system of the present invention's realization.
Preferred embodiment describes in detail
The additional MSE that the cascade compression is introduced can be illustrated in conjunction with shown in Figure 2.It is Q that Fig. 2 (a) shows step-length
1Some reconstruction point and the decision boundaries of uniform quantizer.These reconstruction point are represented with stain, and decision boundaries is represented with short vertical line.N reconstruction point is positioned at Q
n 1(not shown), Q
n 1The decision boundaries of both sides is positioned at D
n 1And D
(n-1) 1(not shown).Decision boundaries roughly is in the centre of two reconstruction point in succession.For a step-length is Q
1Uniform quantizer, reconstruction point is in Q
1The multiple place.An input signal source any drops on two values between the decision boundaries and is quantified as the reconstruction point that is in the middle of two decision boundaries.
Quantizer Q1 is used for the first order (a) of cascade compression, and quantizer Q
2The second level that is used to compress (b).For a single stage compress, that for example use is quantizer Q
2In this embodiment, quantizer Q
2Step-length as shown greater than quantizer Q
1Step-length.It is to be noted that bigger quantization step can reach bigger compression ratio, yet be cost to introduce more loss.Preferably, also be uniform quantizer such as those that in the I of MPEG frame and JPEG compression scheme, use.Yet, also can use other step-lengths and nonuniform quantiza device.
In Fig. 2, x represents the value of the input source signal of requirementization.If x drops on [Q
0 2,
0 2) scope in, with single-stage quantizer (in this case only with level (1)), this value will be quantified as Q
0 2It is to be noted that symbol " [" is illustrated in and comprises this value in the scope, and symbol ") " is illustrated in and does not comprise this value in this scope.Under the situation of two-stage quantizer ((a) and (b)), if x drop on [0, D
0 1) scope in the output of first order quantizer be Q just
0 1, and if it drops on [D
0 1, D
0 2) scope in be Q just
1 1The output of level (a) is then by second quantized level (b), with Q
0 1Be quantified as Q
0 2, and with Q
1 1Be quantified as Q
1 2Therefore, with the cascade compression, [0, D
0 1) scope in the x value will be quantified as Q
0 2, and at [D
0 1, D
0 2) scope in the x value will be quantified as Q
1 2Therefore, compare, adopt the cascade compression will make at scope [D with the single-stage quantizer
0 1, D
0 2) in the x value be subjected to incorrect quantification, thereby have bigger mean square error.Similar, can notice, any at [D
1 1, D
1 2) scope in the x value will be quantified as a Q improperly with the two-stage quantizer
2 2Value, and the single-stage quantizer can be quantified as Q with these values
1 2This additive error is the direct result of cascade compression.
In order to reduce and/or to eliminate this additive error, at first must determine at given quantizer Q
1And Q
2The time can be subjected to the scope of value of the x of incorrect quantification.About this point, consider Q
2Quantizer is at D
n 2The specific decision boundaries at place.Decision boundaries hereto, quantizer Q
1Two immediate decision boundaries, one is to compare D
n 2Big (D
m 1), one is to compare D
n 2Little (D
M-1 1), their position is as with shown in 3 (b).With regard to illustration, consider following situation: situation 1 (Fig. 3 (a)): Q
m 1＜D
n 2And x ∈ [D
n 2, D
m 1)
Single-stage Q
2Quantizer can be quantified as Q with x
N+1 2At two-stage quantizer (Q
1And Q
2) in, Q
1X can be quantified as Q
m 1, and the Q of back
2Quantizer can be quantified as Q with x
n 2Thereby, introduce the additional quantization error.Situation 2 (Fig. 3 (b)): Q
m 1〉=D
n 2And x ∈ [D
N-1 1, D
n 2)
Single-stage Q
2Quantizer can be quantified as Q with x
n 2At two-stage quantizer (Q
1And Q
2) in, Q
1Quantizer can be quantified as Q with x
m 1, and the Q of back
2Quantizer can be quantified as Q with it
(n+1) 2Thereby, introduce the additional quantization error.
For the value of determining x can be subjected to all scopes of incorrect quantification, must be for Q
2All decision boundaries repeat this calculating.Yet, for example under situation, for Q as uniform quantizer used in the JPEG compression
2All decision boundaries only need be 0 to LCM (Q
1, Q
2) scope in carry out aforementioned calculation.LCM (Q
1, Q
2) be two number Q
1And Q
2Least common multiple.The value that this result who calculates can be used for obtaining x be in scope [0, LCM (Q
1, Q
2] outer thereby can be subjected to the scope of incorrect quantification.As can be seen, this calculating is much simple under the situation of uniform quantizer.
For each combination of some pixel values in an environment, the probability distribution of black and white pixel may be different.For example, all be that the probability that will compare a black pixel coder to the probability of a white pixel coder is much bigger in the white environment at one.The probability distribution f (x) of given input source signal (x) just can followingly calculate and use quantizer Q
1And Q
2Cascade quantizes the expectation quantization error of introducing:
E (Q
1, Q
2)=∫ f (x) represents to contain the set that can be subjected to the All Ranges of incorrect quantification just like the value of above definite x for x ∈ ξ symbol ξ.
The calculating of the quantization error of carrying out with following formula is used for selecting the suitable quantizer in the second level of cascade compression.For example, if first order quantizer Q
1, and supposition has two possible quantizer Q
2And Q
2' can be used for the second level, so just can calculate quantization error for these two quantizers.These two possible quantizers are test quantizers.Decide selection quantizer Q by quantizing the error amount minimum
2And Q
2' in which is the most suitable.
If estimate quantizer Q
2Provide the bit rate (bigger speed, less compression) of r and quantization error is E (Q
2)+E (Q
1, Q
2), and quantizer Q
2' estimate the bit rate of r ' is provided and quantization error is E (Q
2')+E (Q
1, Q
2'), so can select the tolerance of quantizer with the conduct recently of bit rate and quantization error.Here, E (Q
2) and E (Q
2') be the quantizer inherent quantization error, quantize caused additive error with cascade and have nothing to do.
Certainly, can use Q discussed above
2Or Q
2' outside quantizer as starting point.Part is according to required total compression selected amount device recently.In selecting initial quantizer, also can adopt trial and error procedure.Can calculate the quantization error of several quantizers as mentioned above like that, select optimal quantizer by this.
As mentioned above, one embodiment of the present of invention relate to the application of adopting JPEG and MPEG compression scheme.These two compression schemes are the piece of 8x8 with the size that the input source data spatially are divided into a series of adjacency, and each piece draws 64 DCT coefficients behind dct transform.Then these DCT coefficients are quantized.What the DC coefficient was carried out is differential coding.63 remaining AC coefficients are encoded to the mode that zero coefficient values is subsequently encoded by the run length of regulation zero coefficient again.
Under the situation of JPEG, the entry of quantization table has determined to be used for the quantification of different DCT coefficients.Can adopt different quantization tables for different zones (for example, brightness and colourity), but quantization table is fixed for a zone.In order in the data recompression that has been subjected to the JPEG compression, to use quantizer system of selection of the present invention, must know the probability distribution f (x) of each DCT coefficient.Empirical discovery, the distribution of AC DCT coefficient are deferred to Laplacian and are distributed.It is to be noted that the parameter related with the Laplacian distribution is different for different DCT coefficients.This parameter can be estimated, perhaps draws a different distribution according to available packed data, such as Rayleigh or Gaussian distribution.
Be appreciated that for DCT coded image and MPEG error term and can distribute with Laplacian.The MPEG compression scheme is encoded to error term and pictorial information with DCT.Error term obtains according to the MPEG movement compensating algorithm.An error term is to obtain by after the piece on another pattern of sequence deducts an image block difference being carried out DCT from one.This makes that pattern has only at image under the situation of a small amount of change can be with less bits of encoded.
In addition, for the situation of MPEG, a preferred embodiment at be I (intraframe coding) frame in the mpeg format.The I frame includes only not the Intra-coded blocks with reference to other patterns.These frames can be as the access point at random in the sequence.Also can use other mpeg frame types such as P (predictive coding) frame and B (two-way interpolation) frame.
For example, for each (I) frame in the MPEG video, quantized value depends on quantification scale (quantizer_scale) and quantization table.Can be with different quantization tables for colourity and brightness.Quantization table is fixed for each frame, but quantizing scale can change for each macro block.In one embodiment, select the quantification scale of each mpeg frame macro block with the quantizer back-and-forth method of above argumentation.Yet, it is to be noted, for each DCT coefficient, can not change the quantification scale.This value is fixed for whole macro block.Because the human visual system is relatively more responsive to the quantization error of low frequency coefficient, therefore preferable is to be chosen to make the average quantization error of selected low frequency coefficient to reduce to minimum with quantizing scale.
Fig. 4 shows one can realize video treatment system 20 of the present invention.For example, system 20 can represent a television set, a set-top box, desk-top, an above-knee or palmtop computer, personal digital assistant (PDA) one by one, video storage device such as video cassette recorder (VCR), digital VTR (DVR), TiVO equipment, and the some parts or the combination of these and other equipment.System 20 comprises one or more video/image source 22,24, one processors 25 of one or more input/output units, and a memory 26.Video/image source 22 can be represented for example television receiver, a VCR, perhaps other video storage devices.Source 22 also can represent one or more by for example one connect such as the some parts of the network of the Internet, wide area network, metropolitan area network, local area network (LAN), terrestrial broadcast systems, cable television network, satellite network, wireless network or telephone network and these and other type or combination network from server receiver, video/image.
Input/output unit 24, processor 25 and memory 26 are communicated by letter by communication medium 27.Communication medium 27 can be represented for example bus, a communication network, and the inside of one or more circuit, circuit card or other equipment connects, and the some parts and the combination of these and other communication medium.The video of 22 inputs is handled according to the one or more software programs by processor 25 execution that are stored in the memory 26 from the source, produces output video/image, offers a display unit 28 such as television indicator, computer monitor.
In a preferred embodiment, the expectation quantification sum of errors that calculates the cascade compression selects suitable quantizer to be realized by the computer-readable code that system 20 carries out.This code can be stored in the memory 26, perhaps reads/downloads from a medium such as CD-ROM or floppy disk.In other embodiments, can replace software instruction or combine the realization invention with software instruction with hardware circuit.
The concrete configuration that is appreciated that system 20 as shown in Figure 4 is just as an example.Those skilled in the art can recognize that the present invention can realize with multiple alternative system configuration.
Though abovely describe the present invention, be appreciated that this is not these embodiment that the present invention just is confined to here to be disclosed with regard to some specific embodiments.For example, the present invention is not limited to any concrete compression scheme, frame type or probability distribution.On the contrary, the present invention has covered various structures and the modification in the spirit and scope that are included in appended claims.
Claims (17)
1. method that is used for a cascade compressibility (12), described method comprises the following steps:
Determine by one second of described cascade compressibility or the expectation quantization error introduced of back level;
To at least two can be used for second or the expectation quantization error of the quantizer of back level compare; And
From described two quantizers, select a quantizer at least according to comparative result.
2. in accordance with the method for claim 1, wherein said expectation quantization error is determined with a probability-distribution function.
3. in accordance with the method for claim 2, wherein said probability-distribution function is that a Laplacian distributes.
4. in accordance with the method for claim 2, wherein said selection step is selected a quantizer according to the expectation bit rate and the ratio of expectation quantization error.
5. in accordance with the method for claim 2, wherein said probability-distribution function is determined according to the input data (10) of need compression.
6. in accordance with the method for claim 1, wherein said at least two quantizers are determined according to the required compression ratio of described cascade compressibility.
7. in accordance with the method for claim 1, wherein said cascade compressibility comprises one of JPEG and mpeg system at least.
8. in accordance with the method for claim 1, wherein said determining step comprises that the value of determining input data will be by the set of the scope of the incorrect quantification of described cascade compressibility.
9. medium that contains the code that is useful on a cascade compressibility (12), described code comprises:
Calculating is by one second or definite code of the back grade expectation quantization error of introducing of described cascade compressibility;
Can be used for second or the comparison code that compares of the expectation quantization error of the quantizer of back level at least two; And
The result who draws according to code relatively considers to select the selection code of a quantizer from described at least two quantizers.
10. according to the described medium of claim 9, wherein said expectation quantization error is determined with a probability-distribution function.
11. according to the described medium of claim 9, wherein said probability-distribution function is that a Laplacian distributes.
12. according to the described medium of claim 9, wherein said cascade compressibility comprises one of JPEG and mpeg system at least.
13. a cascade compression device (20), described equipment comprises:
The memory of a store executable code (26); And
An execution is stored in the processor (25) of the code in the described memory (26), described processor is used for: (i) determine by one second of described cascade compressibility or the expectation quantization error introduced of back level, (ii) to can be used for second or the expectation quantization error of at least two quantizers of back level compare, and (iii) can from described two quantizers, select a quantizer at least according to comparative result.
14. according to the described equipment of claim 13 (20), wherein said expectation quantization error is determined with a probability-distribution function.
15. according to the described equipment of claim 14 (20), wherein said probability-distribution function is that a Laplacian distributes.
16. according to the described equipment of claim 13 (20), wherein said cascade compressibility (12) comprises one of JPEG and mpeg system at least.
17. a cascade compressibility (12), described cascade compressibility comprises:
Determine device (25,26) by the expectation quantization error of described cascade compressibility introducing;
To can be used for one second of described cascade compressibility or the device (25,26) that compares of the definite described expectation quantization error of possible quantizer of back level according to some; And
May select the device (25,26) of a quantizer the quantizer from described to reduce described expectation quantization error.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US43350799A | 1999-11-03 | 1999-11-03 | |
| US09/433,507 | 1999-11-03 |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1342369A true CN1342369A (en) | 2002-03-27 |
| CN1186940C CN1186940C (en) | 2005-01-26 |
Family
ID=23720384
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB008045496A Expired - Fee Related CN1186940C (en) | 1999-11-03 | 2000-10-13 | Improved cascaded compression method and system for digital video and images |
Country Status (5)
| Country | Link |
|---|---|
| EP (1) | EP1145562A1 (en) |
| JP (1) | JP2003513563A (en) |
| KR (1) | KR100744442B1 (en) |
| CN (1) | CN1186940C (en) |
| WO (1) | WO2001033860A1 (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101073272B (en) * | 2004-10-20 | 2015-04-29 | 皇家飞利浦电子股份有限公司 | Method of reducing quantization noise |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE69224227T2 (en) * | 1991-04-18 | 1998-07-09 | Ampex | Method and arrangement for determining a quantization factor for processes with data compression and decompression |
| FR2743246B1 (en) * | 1995-12-29 | 1998-01-23 | Thomson Broadcast Systems | METHOD AND DEVICE FOR COMPRESSING DIGITAL DATA |
| JP2001518247A (en) * | 1997-02-25 | 2001-10-09 | ブリティッシュ ブロードキャスティング コーポレーション | Digital signal compression encoding method using improved quantization processing |
-
2000
- 2000-10-13 KR KR1020017008507A patent/KR100744442B1/en not_active IP Right Cessation
- 2000-10-13 EP EP00967876A patent/EP1145562A1/en not_active Withdrawn
- 2000-10-13 CN CNB008045496A patent/CN1186940C/en not_active Expired - Fee Related
- 2000-10-13 JP JP2001534882A patent/JP2003513563A/en not_active Withdrawn
- 2000-10-13 WO PCT/EP2000/010158 patent/WO2001033860A1/en active Application Filing
Also Published As
| Publication number | Publication date |
|---|---|
| EP1145562A1 (en) | 2001-10-17 |
| WO2001033860A1 (en) | 2001-05-10 |
| CN1186940C (en) | 2005-01-26 |
| KR100744442B1 (en) | 2007-08-01 |
| JP2003513563A (en) | 2003-04-08 |
| KR20010089765A (en) | 2001-10-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1078428C (en) | Image coding method and apparatus therefor | |
| US7843998B2 (en) | Method for improved entropy coding | |
| EP1529401B1 (en) | System and method for rate-distortion optimized data partitioning for video coding using backward adaptation | |
| US7003034B2 (en) | Fine granularity scalability encoding/decoding apparatus and method | |
| CN1164167A (en) | Adaptive quantizer for use in video signal encoding system | |
| CN1347621A (en) | Reducing 'blocking picture' effects | |
| CN1155788A (en) | Method and device for compressing digital data | |
| US20070041653A1 (en) | System and method of quantization | |
| CN1813480A (en) | Early detection of zeros in the transform domain | |
| CN1315109A (en) | Image processing circuit and method for modifying a pixel valve | |
| US5822000A (en) | Video encoding using rearrangement of transform coefficients and inter-block correlation | |
| EP2196031B1 (en) | Method for alternating entropy coding | |
| US5724096A (en) | Video signal encoding method and apparatus employing inter-block redundancies | |
| CN1186940C (en) | Improved cascaded compression method and system for digital video and images | |
| EP1014725A1 (en) | Decoder and decoding method | |
| US20050157790A1 (en) | Apparatus and mehtod of coding moving picture | |
| CN1112336A (en) | Image signal decoding apparatus having an encoding error compensation | |
| US5825422A (en) | Method and apparatus for encoding a video signal based on inter-block redundancies | |
| Jaksic et al. | Analisys brightness effect on quality pictures after compression with JPEG and SPIHT compression method | |
| Burg | Image and video compression: the principles behind the technology | |
| KR100188680B1 (en) | Apparatus and method of selecting the dct mode | |
| Dhavale et al. | Discrete cosine transform based image compression | |
| KR100561164B1 (en) | DCT coefficient requantization method | |
| Nanavati et al. | Wavelets: Applications to image compression-ii | |
| Leung et al. | Compression techniques for still image and motion video |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20050126 Termination date: 20091113 |