CN1595453A - Image compression method based on wavelet transformation - Google Patents
Image compression method based on wavelet transformation Download PDFInfo
- Publication number
- CN1595453A CN1595453A CN 200410023324 CN200410023324A CN1595453A CN 1595453 A CN1595453 A CN 1595453A CN 200410023324 CN200410023324 CN 200410023324 CN 200410023324 A CN200410023324 A CN 200410023324A CN 1595453 A CN1595453 A CN 1595453A
- Authority
- CN
- China
- Prior art keywords
- coefficient
- wavelet
- conspicuousness
- divided
- wavelet coefficient
- 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
- 238000000034 method Methods 0.000 title claims abstract description 51
- 230000009466 transformation Effects 0.000 title claims abstract description 15
- 230000006835 compression Effects 0.000 title abstract description 9
- 238000007906 compression Methods 0.000 title abstract description 9
- 238000013139 quantization Methods 0.000 claims abstract description 15
- 238000009826 distribution Methods 0.000 claims abstract description 6
- 238000013461 design Methods 0.000 claims description 3
- 230000003247 decreasing effect Effects 0.000 abstract 1
- 230000002093 peripheral effect Effects 0.000 abstract 1
- 238000007670 refining Methods 0.000 abstract 1
- 238000010586 diagram Methods 0.000 description 5
- 238000012360 testing method Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 101000888552 Xenopus laevis CLIP-associating protein 1-B Proteins 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 230000002950 deficient Effects 0.000 description 2
- 238000013178 mathematical model Methods 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 235000002566 Capsicum Nutrition 0.000 description 1
- 241001504519 Papio ursinus Species 0.000 description 1
- 239000006002 Pepper Substances 0.000 description 1
- 235000016761 Piper aduncum Nutrition 0.000 description 1
- 235000017804 Piper guineense Nutrition 0.000 description 1
- 244000203593 Piper nigrum Species 0.000 description 1
- 235000008184 Piper nigrum Nutrition 0.000 description 1
- 230000000903 blocking effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 230000006735 deficit Effects 0.000 description 1
- 238000009792 diffusion process Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 239000004744 fabric Substances 0.000 description 1
- 239000012467 final product Substances 0.000 description 1
- 239000010978 jasper Substances 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000003908 quality control method Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
- 230000017105 transposition Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Abstract
The invention relates to a wavelet transformation image compression method, which is composed of the following steps. First, do the wavelet transformation to the digital image to get the wavelet coefficients. Second, each wavelet coefficient is divided by a quantization coefficient. Third, divide the wavelet coefficients into blocks according to its subband position. Fourth, scan every coefficient block along the level. During the scanning, each wavelet coefficient is classified into 13 types according to the significance of the peripheral wavelet coefficients, the sign of the coefficients, the refining and general output. The arithmetic coding is used according to the different distribution model. The invention may cancel the OT procedure, and decrease the needed number of the operating period and the coding system delay. The relative stability of the image quality is guaranteed during the coding procedure. And the coding symbol number is decreased to reduce the pressure on the binary arithmetic encoder.
Description
Technical field:
The invention belongs to areas of information technology, be specifically related to a kind of digital image compression coding method based on wavelet transformation.
Background technology:
At present, JPEG2000 standard code method is all adopted in the digital image compression coding method of wavelet transformation, EBCOT coding flow process in the JPEG2000 standard is as follows: as shown in Figure 1, it is to obtain wavelet coefficient after the conversion of original image process multilevel wavelet, then the wavelet coefficient of each subband is divided into one by one coefficient block (JPEG2000 suggestion be divided into 64 * 64 or 32 * 32 for well), each piece carries out importance priority encoding (EBC) earlier, realizes that again global optimization blocks (OT).In order to realize the importance priority encoding, each piece wavelet coefficient used former representation after, most significant digit is-symbol position (0 for just, 1 for negative), other are the absolute value of coefficient.The set of the same position in the coefficient absolute value is called plane (as shown in Figure 2), the plane scanning one by one from an absolute value high position towards low plane in the cataloged procedure.To each coefficient in the piece, if its value in plane k is 1, and all be 0 in the plane higher than k, claim that then this coefficient is significant in plane k..0.Scanning is three times during each bit plane coding, scanning for the first time to around have and encode at the significant coefficient of the plane higher than current plane, be called conspicuousness diffusion (Significance propagation), scanning for the second time is the present bit face amount of the significant coefficient of output, be called refinement (Magnitude refinement), scanning for the third time is not have detected remarkable coefficient to scan to scanning for the first time, is called cleaning (Cleanup).Like this, be considered to the just preferential output of important information wavelet coefficient information, and the information of each scanning process output of each plane in the code stream can block all, this coding characteristic is called " embedded encoded ".
The concrete cataloged procedure of JPEG2000 as shown in Figure 3, after each coefficient block coding is finished, again the coding result of each coefficient block is blocked (OT) by optimization and combine, constitute so-called quality layers, can realize the importance priority encoding of entire image like this.But do like this and increased the needed clock periodicity of coding greatly, and the theory of using OT of overall importance to increase coded system postpones, in addition, though this way can accurately be controlled the code check of each independent image, there is not effective method that the image sequence quality of a plurality of image constructions is carried out stable control.
The EBCOT scrambler is the core of JPEG2000 scrambler, because each plane is successively encoded, and each plane needs three scanning, caused cataloged procedure very consuming time, particularly in the hardware implementation procedure, the periodicity all the time that takies is many especially, considerably beyond wavelet transformation part and arithmetic coding part, becomes the bottleneck of whole JPEG2000 coded system.
Summary of the invention:
Technical matters to be solved by this invention is the defective that overcomes above-mentioned prior art, a kind of needed clock periodicity of coding that reduces is provided, simultaneously can to code check and quality be controlled and can realize the digital image compression coding method of the wavelet transformation of higher pixel rate.
The present invention solves the problems of the technologies described above by following technical scheme.It comprises the steps: 1) digital picture is carried out wavelet transformation obtain wavelet coefficient; 2) to each wavelet coefficient divided by a quantization parameter, this quantization parameter is between 1.0 to 1000.0; 3) wavelet coefficient is carried out piecemeal according to the subband position at place; 4) each coefficient block scans by plane, in the plane scanning process each wavelet coefficient is divided into 13 classes according to conspicuousness, wavelet coefficient symbol, refinement and the general output of wavelet coefficient on every side, wherein conspicuousness is divided into 6 classes, classification number is 0-5, the wavelet coefficient symbol falls into 5 types, and classification number is 6-10, and the classification number of refinement is 11, general output category number is 12, uses arithmetic encoder to encode according to type according to different distributed models then.Described step 3) is that wavelet coefficient is divided into the coefficient block that is no more than 32 * 32 sizes, subband is not striden on the border of coefficient block, each 32 * 32 coefficient block is divided into 16 * 16 the sub-piece of coefficient again, and each coefficient block is divided into four types of LL, LH, HL and HH according to the subband position at place; Described step 4) is: 1. use the highest significant position face number also output of classification number 12 design factor pieces; 2. use classification number 12 that the sub-piece of each coefficient of 16 * 16 is calculated number also output of its highest significant position face; 3. begin to scan in the face of coefficient block from the highest significant position face, skip the sub-piece of coefficient in the scanning process above highest significant position face number to lowest order; 4. in scanning process according to around the conspicuousness situation of coefficient the conspicuousness characteristic distribution of the inapparent coefficient of current plane is classified, then the conspicuousness of coefficient is delivered to the output of arithmetic encoder coding according to classification results.In scanning process to significantly coefficient according to around the conspicuousness situation of wavelet coefficient classify and comprise the steps: 1. earlier according to the conspicuousness situation around the coefficient, calculate the weights of this coefficient, then the weighting table according to LL, LH, HL and HH type blocks adds this weights if significantly; 2. weights adjustment is obtained 6 kinds of conspicuousness classification numbers; 3. to coefficient significantly, according to the value of classification number 11 these coefficients of output at current plane; 4. to the coefficient of unknown conspicuousness, the actual conspicuousness of coefficient is exported by the classification number that obtains according to the conspicuousness classification number; If this coefficient is remarkable really, the sign bit of output coefficient.
The present invention has following technique effect:
1) the present invention proposes a kind of coding method of a plane single pass, this method is better than EBCOT on compression performance, simpler than EBCOT aspect realizability, cancelled the OT process, when realizing, hardware can significantly reduce the needed periodicity of operation, and can realize the hardware pipeline operation, reduced the delay of coded system.
2) the present invention divides time-like to adopt the bigger mathematical model in relevant range carrying out coefficient, each coefficient is classified according to the state of 24 coefficients around it, considered the Long-Range Correlation of wavelet coefficient than the scheme of 8 coefficients around the JPEG2000 basis more fully.Handle like this, can guarantee to be not less than JPEG2000 in single pass lower compression performance.
3) the present invention has adopted the notion of the sub-piece of coefficient, further reduced the coded identification number, alleviated the pressure of binary arithmetic encoder greatly, make the symbolic number and the ratio of visual sample number be issued to 1.5: 1, under the situation of a cover logic working, can realize higher pixel rate in typical quality situation.
4) the present invention is by carrying out scalar quantization to wavelet coefficient, can realize the control of quality control and code check, guarantee the relative stability of image sequence picture quality in cataloged procedure, avoided the phenomenon of picture quality big rise and big fall, be particularly suitable for the coding of the preferential digital video of quality.
Description of drawings:
Fig. 1 is a wavelet coefficient piecemeal synoptic diagram;
Fig. 2 is the plane of Wavelet Coefficient Blocks;
Fig. 3 is the picture coding process synoptic diagram of prior art JPEG2000;
Fig. 4 is a picture coding process synoptic diagram of the present invention;
Fig. 5 is divided into the sub-piece synoptic diagram of wavelet coefficient for Wavelet Coefficient Blocks of the present invention;
Fig. 6 is the scanning sequency of same plane in the sub-piece of wavelet coefficient of the present invention;
Fig. 7 is the present invention's process flow diagram of encoding.
Embodiment:
The applicant is through research and use discovery, JPEG2000 is embedded encoded in order to realize in the prior art, in to the wavelet coefficient cataloged procedure, adopted plane scanning one by one, each plane carries out the way of three scannings again and carries out block encoding (EBC), in the scope of image block, the coefficient block behind the coding is optimized again after all the coefficient block coding is finished and blocks (OT), this method can increase the needed clock periodicity of coding greatly, and the coefficient block behind the coding is optimized the theory of blocking (OT) and increased coded system postpones, in addition, though this way can accurately be controlled the code check of each independent image, there is not effective method that the image sequence quality of a plurality of image constructions is carried out stable control.
The present invention is directed to above-mentioned defective, a kind of coding method of a plane single pass has been proposed, this method can be cancelled the OT process, when realizing, hardware can significantly reduce the needed periodicity of operation, reduced the delay of coded system, by wavelet coefficient is carried out scalar quantization, can accurately control code check equally, also can guarantee the relative stability of image sequence picture quality in cataloged procedure simultaneously, wavelet coefficient after quantizing is carried out carrying out sub-piecemeal again behind the piecemeal, can reduce the coded identification number, alleviate the pressure of binary arithmetic encoder greatly.
The present invention is described in detail in detail below.
Concrete implementation step of the present invention is as follows:
1, original image is carried out the multilevel two-dimensional wavelet transform, use W97-2 wavelet filter in the prior art, as shown in Figure 1, can obtain the wavelet coefficient of each subband, be divided into LL, LH according to the position, four types of HL and HH, when decomposed class was N, sub band number was 3N+1;
2, obtain wavelet coefficient by wavelet transformation after, wavelet coefficient is carried out scalar quantization, promptly to each wavelet coefficient divided by a quantization parameter, this quantization parameter is generally between 1.0 to 1000.0, quantization parameter is big more, and image impairment is big more, and the code stream of output is just more little.This scalar quantization can be specified different quantization parameters to different subbands according to the requirement of application system, also can only specify unified quantization parameter.Quantization parameter is recorded in the output code flow, uses for decoding;
3, the wavelet coefficient after quantizing carries out piecemeal according to method shown in Figure 1, notices that piece do not stride the subband border, and block size is 32 * 32, if not enough 32 * 32, then according to actual size definition Wavelet Coefficient Blocks.Each coefficient block is divided into four types of LL, LH, HL and HH according to the subband position at place.The processing sequence of Wavelet Coefficient Blocks is: according to the high frequency resolution of resolution from LL to the outermost, in the same resolution according to LH, HL, the order of HH subband, in the same subband according to from left to right, partitioning scheme from top to bottom again;
4, each coefficient block scans by plane, in the plane scanning process each wavelet coefficient is divided into 13 classes according to conspicuousness, wavelet coefficient symbol, refinement and the general output of wavelet coefficient on every side, wherein conspicuousness is divided into 6 classes, classification number is 0-5, the wavelet coefficient symbol falls into 5 types, and classification number is 6-10, and the classification number of refinement is 11, general output category number is 12, uses arithmetic encoder to encode according to type according to different distributed models then.The present invention divides time-like to adopt the bigger mathematical model in relevant range carrying out wavelet coefficient, carry out the correlativity statistics by wavelet coefficient to great amount of images, we find that the correlativity of wavelet coefficient is distant, so the present invention classifies according to the state of 24 coefficients around it to each coefficient, considered the Long-Range Correlation of wavelet coefficient more fully than the scheme of 8 coefficients around the JPEG2000 basis.Handle like this, can guarantee to be not less than JPEG2000 in single pass lower compression performance.The design of concrete sorting criterion has then taken into full account the distribution characteristics of wavelet coefficient, has considered the complicacy that realizes again, draws according to actual compression result's statistics of great amount of images.
As shown in Figure 5, the present invention presses classification number 12 to Wavelet Coefficient Blocks and calculates highest significant position faces number, each Wavelet Coefficient Blocks is divided into 4 16 * 16 the sub-piece of wavelet coefficient again, and 4 sub-pieces of wavelet coefficient are calculated highest significant position face number in the sub-piece.Then the highest significant position face number (0..15) of Wavelet Coefficient Blocks is outputed in the arithmetic encoder (from MSB to LSB totally 4) by classification number 12, wherein 15 these Wavelet Coefficient Blocks of expression do not have non-zero wavelet coefficient.If this piece does not have non-zero wavelet coefficient, then directly enter next Wavelet Coefficient Blocks.The sub-piece of each wavelet coefficient uses the highest significant position face number of the sub-piece of classification number 12 outputs;
5, begin to carry out plane scanning one by one to lowest order in the face of Wavelet Coefficient Blocks from the highest significant position face, the scanning sequency of same plane is carried out according to the sub-block number of wavelet coefficient.Scan according to the described scanning sequency of Fig. 6 in the sub-piece of wavelet coefficient inside, whether skip this sub-piece according to the highest significant position face number decision of the sub-piece of this wavelet coefficient.Each Wavelet Coefficient Blocks is an absolute coding in the scanning process, but is to share conspicuousness between 4 sub-pieces of wavelet coefficient of a Wavelet Coefficient Blocks;
6, in scanning process, at higher plane wavelet coefficient significantly, according to classification number 11 these wavelet coefficients of output in the value of this plane in arithmetic encoder; To in higher plane, also there not being significant wavelet coefficient, classify according to the known remarkable situation of coefficient on every side, export the conspicuousness at this plane (0-is not remarkable, and 1-is remarkable) of this wavelet coefficient then according to classification number, its concrete scheme of classification is as follows:
A. earlier according to the conspicuousness situation around the coefficient, calculate the weights of this coefficient, according to following weighting table, if significantly, then add these weights:
The weight table of LL and LH type blocks is as follows:
| ????-2 | ????-1 | ????0 | ????1 | ????2 | |
| ????-2 | ????1 | ????2 | ????4 | ????2 | ????1 |
| ????-1 | ????4 | ????8 | ????16 | ????8 | ????4 |
| ????0 | ????8 | ????32 | ????X | ????32 | ????8 |
| ????1 | ????4 | ????8 | ????16 | ????8 | ????4 |
| ????2 | ????1 | ????2 | ????4 | ????2 | ????1 |
The weight table of HL type blocks is the transposition of LH weight table;
The weight table of HH type blocks:
| ????-2 | ????-1 | ????0 | ????1 | ????2 | |
| ????-2 | ????2 | ????2 | ????2 | ????2 | ????2 |
| ????-1 | ????2 | ????16 | ????16 | ????16 | ????2 |
| ????0 | ????2 | ????16 | ????X | ????16 | ????2 |
| ????1 | ????2 | ????16 | ????16 | ????16 | ????2 |
| ????2 | ????2 | ????2 | ????2 | ????2 | ????2 |
Wherein, the current coefficient of " X " expression position.Note a borderline some computing block internal state (it is not remarkable that piece is all thought) in the computation process outward.
B. weights adjustment is obtained classification number
According to the form below adjustment obtains classification number:
| Weights | ??＞＝72 | ????[48，71] | ????[24，47] | ????[16，23] | ????[8，15] | ????[0，7] |
| Classification | ??5 | ????4 | ????3 | ????2 | ????1 | ????0 |
To the remarkable coefficient in the scanning process, if in this plane significantly, then export the symbol of this wavelet coefficient, algorithm is identical with SC algorithm among the JPEG2000.Concrete arthmetic statement is as follows:
Sign bit is to predict according to the situation of 4 coefficients around the coefficient, and sign bit is fallen into 5 types.
The symbol of coefficient be considered to 8 coefficients on every side in horizontal neighbours and vertical neighbors to amount to four coefficients relevant.Each adjacent coefficient has three kinds of states: significantly, not remarkable significantly for just for negative, so have 81 kinds of combinations.In order to reduce number of combinations, we contribute and vertical contribution according to the following table calculated level with two vertical adjacent coefficients two horizontal adjacent coefficients earlier according to symmetry:
Have 9 kinds of combinations then, again according to 5 kinds of combinations of following table boil down to, each makes up the classification number as a symbol:
| The level contribution | Vertical contribution | Classification number | ????XORbit |
| ????1 | ????1 | ????10 | ????0 |
| ????1 | ????0 | ????9 | ????0 |
| ????1 | ????-1 | ????8 | ????0 |
| ????0 | ????1 | ????7 | ????0 |
| ????0 | ????0 | ????6 | ????0 |
| ????0 | ????-1 | ????7 | ????1 |
| ????-1 | ????1 | ????8 | ????1 |
| ????-1 | ????0 | ????9 | ????1 |
| ????-1 | ????-1 | ????10 | ????1 |
During output symbol, if canonical note sign bit is 0, remember that then sign bit is 1 if bear, output (sign bit) xor (XORbit) gets final product then.
7, use arithmetic encoder to encode according to the wavelet coefficient classification by the different distributions model.The arithmetic encoder identical (logical symbol is seen Fig. 7) that uses among this arithmetic encoder and the JPEG2000, concrete algorithm flow is identical with the arithmetic entropy coding of ISO/IEC FCD15444-1 appendix C with the concrete distribution probability parameter of using.But use the init state parameter and the JPEG2000 of distributed model different among the present invention.These 13 distributed models to have related parameter to gather as follows:
| Sequence number | The user | Number of categories | Classification number | The original state of arithmetic entropy coding device (index, MPS) |
| ??1 | The conspicuousness classification | ????6 | ????0..5 | Wherein use (4,0) No. 0, No. 1 use (3,0), other uses (0,0) |
| ??2 | The wavelet coefficient symbol | ????5 | ????6..10 | ??(0，0) |
| ??3 | Refinement | ????1 | ????11 | ??(0，0) |
| ??4 | General output | ????1 | ????12 | ??(46，0) |
| Add up to | ------------------- | ???13 | ????---- |
＜performance relatively 〉
Comparison when following form is coding method of the present invention and JPEG2000 use CDF97 wavelet filter.
1, test way: to the test pattern of some width of cloth current international practices, use the JPEG2000 coded program that software of the present invention is realized and the JPEG tissue is issued (the Standard test programme VM6.1 of JPEG issue) to encode with this typical code check parameter, behind the picture decoding of two kinds of algorithm codings, use the snr computation program (imgcmp of JPEG tissue suggestion, ISO/IEC N2415 JasPer V1.700.0) carries out PSNR with former figure and calculate, obtain PSNR result.
2, test result
Table 1 JPEG2000 and PSNR result of the present invention are relatively
| Image Name | Size of the present invention (byte) | PSNR of the present invention (dB) | JPEG2000 size (byte) | ??JPEG2000 ??PSNR(dB) |
| ?Lena | ????8011 | ????34.01 | ????8012 | ????33.95 |
| ?Barbara | ????11079 | ????29.86 | ????11075 | ????29.88 |
| ?baboon | ????14977 | ????25.28 | ????14949 | ????25.14 |
| ?girl512 | ????6967 | ????34.99 | ????6968 | ????34.88 |
| ?pepper | ????7789 | ????33.92 | ????7787 | ????33.98 |
| ?Jet | ????8907 | ????32.78 | ????8885 | ????32.76 |
| ?couple | ????6863 | ????38.97 | ????6836 | ????39.06 |
| ?woman | ????170689 | ????30.18 | ????170721 | ????30.14 |
| ?bike | ????195401 | ????30.33 | ????195389 | ????30.42 |
| ?cafe | ????312045 | ????26.46 | ????312069 | ????26.35 |
Claims (4)
1, a kind of method for compressing image based on wavelet transformation, it comprises the steps:
1) digital picture is carried out wavelet transformation and obtain wavelet coefficient;
2) to each wavelet coefficient divided by a quantization parameter, this quantization parameter is between 1.0 to 1000.0;
3) wavelet coefficient is carried out piecemeal according to the subband position at place;
4) each coefficient block scans by plane, in the plane scanning process each wavelet coefficient is divided into 13 classes according to conspicuousness, wavelet coefficient symbol, refinement and the general output of wavelet coefficient on every side, wherein conspicuousness is divided into 6 classes, classification number is 0-5, the wavelet coefficient symbol falls into 5 types, and classification number is 6-10, and the classification number of refinement is 11, general output category number is 12, uses arithmetic encoder to encode according to type according to different distributed models then.
2, the method for compressing image based on wavelet transformation according to claim 1, it is characterized in that described step 3) is that wavelet coefficient is divided into the coefficient block that is no more than 32 * 32 sizes, subband is not striden on the border of coefficient block, each 32 * 32 coefficient block is divided into 16 * 16 the sub-piece of coefficient again, and each coefficient block is divided into four types of LL, LH, HL and HH according to the subband position at place.
3, the method for compressing image based on wavelet transformation according to claim 1 is characterized in that described step 4) is:
1. use the highest significant position face number also output of classification number 12 design factor pieces;
2. use classification number 12 that the sub-piece of each coefficient of 16 * 16 is calculated number also output of its highest significant position face;
3. begin to scan in the face of coefficient block from the highest significant position face, skip the sub-piece of coefficient in the scanning process above highest significant position face number to lowest order;
4. in scanning process according to around the conspicuousness situation of coefficient the conspicuousness characteristic distribution of the inapparent coefficient of current plane is classified, then the conspicuousness of coefficient is delivered to the output of arithmetic encoder coding according to classification results.
4, according to the method for compressing image described in the claim 3 based on wavelet transformation, it is characterized in that in scanning process to significantly coefficient according to around the conspicuousness situation of wavelet coefficient classify and comprise the steps:
1. earlier according to the conspicuousness situation around the coefficient, calculate the weights of this coefficient, if significantly, then the weighting table according to LL, LH, HL and HH type blocks adds this weights;
2. weights adjustment is obtained 6 kinds of conspicuousness classification numbers;
3. to coefficient significantly, according to the value of classification number 11 these coefficients of output at current plane;
4. to the coefficient of unknown conspicuousness, the actual conspicuousness of coefficient is exported by the classification number that obtains according to the conspicuousness classification number; If this coefficient is remarkable really, the sign bit of output coefficient.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200410023324 CN1276391C (en) | 2004-06-18 | 2004-06-18 | Image compression method based on wavelet transformation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200410023324 CN1276391C (en) | 2004-06-18 | 2004-06-18 | Image compression method based on wavelet transformation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1595453A true CN1595453A (en) | 2005-03-16 |
| CN1276391C CN1276391C (en) | 2006-09-20 |
Family
ID=34663402
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 200410023324 Expired - Fee Related CN1276391C (en) | 2004-06-18 | 2004-06-18 | Image compression method based on wavelet transformation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1276391C (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100409693C (en) * | 2006-07-13 | 2008-08-06 | 王国秋 | Orthogonal transformation method for image and video compression |
| CN102158701A (en) * | 2011-04-19 | 2011-08-17 | 湖南大学 | Compressed sensing theory-based classification quantification image coding method |
| CN101079635B (en) * | 2007-05-24 | 2011-10-05 | 北京邮电大学 | Joint scalar quantity quantification method and level quantification method of self-adapted scalar quantity adjustment |
| CN102316324A (en) * | 2011-08-24 | 2012-01-11 | 北京航空航天大学 | Image coding prediction method based on local minimum entropy |
| CN106664408A (en) * | 2014-06-04 | 2017-05-10 | 简·克劳德·科林 | Adaptive precision and quantification of a wavelet transformed matrix |
-
2004
- 2004-06-18 CN CN 200410023324 patent/CN1276391C/en not_active Expired - Fee Related
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100409693C (en) * | 2006-07-13 | 2008-08-06 | 王国秋 | Orthogonal transformation method for image and video compression |
| CN101079635B (en) * | 2007-05-24 | 2011-10-05 | 北京邮电大学 | Joint scalar quantity quantification method and level quantification method of self-adapted scalar quantity adjustment |
| CN102158701A (en) * | 2011-04-19 | 2011-08-17 | 湖南大学 | Compressed sensing theory-based classification quantification image coding method |
| CN102158701B (en) * | 2011-04-19 | 2012-07-25 | 湖南大学 | Compressed sensing theory-based classification quantification image coding method |
| CN102316324A (en) * | 2011-08-24 | 2012-01-11 | 北京航空航天大学 | Image coding prediction method based on local minimum entropy |
| CN102316324B (en) * | 2011-08-24 | 2013-08-21 | 北京航空航天大学 | Image coding prediction method based on local minimum entropy |
| CN106664408A (en) * | 2014-06-04 | 2017-05-10 | 简·克劳德·科林 | Adaptive precision and quantification of a wavelet transformed matrix |
| CN106664408B (en) * | 2014-06-04 | 2019-08-20 | 简·克劳德·科林 | A method of compression digital picture |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1276391C (en) | 2006-09-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1181690C (en) | Encoding method for compression of video sequence | |
| CN1230786C (en) | Embedded quadtree wavelets in image compression | |
| WO1999017257A2 (en) | System and method for compressing images using multi-threshold wavelet coding | |
| US20080008395A1 (en) | Image compression based on union of DCT and wavelet transform | |
| CN101075348A (en) | Method and device for compressing image | |
| CN103581691B (en) | A kind of towards sparse coefficient efficiently can parallel image coding method | |
| George et al. | Image compression based on wavelet, polynomial and Quadtree | |
| CN1276391C (en) | Image compression method based on wavelet transformation | |
| CN102790882B (en) | Encoding method of remote sensing image | |
| Pan et al. | Context-based embedded image compression using binary wavelet transform | |
| CN102307303B (en) | Ternary-representation-based image predictive coding method | |
| Zhu et al. | An improved SPIHT algorithm based on wavelet coefficient blocks for image coding | |
| Yuan et al. | Novel embedded image coding algorithms based on wavelet difference reduction | |
| CN109951711A (en) | A kind of EZW data compression method based on random adjusting thresholds | |
| CN102316324B (en) | Image coding prediction method based on local minimum entropy | |
| Ouafi et al. | A modified embedded zerotree wavelet (MEZW) algorithm for image compression | |
| CN1571514A (en) | An embedded image compression technique based on wavelet transformation | |
| CN1822050A (en) | Full phase anticosine double orthogonal transformation and its improving method for JPEG | |
| CN1637782A (en) | Quad tree image compressing and decompressing method based on wavelet conversion prediction | |
| Zhu et al. | Embedded Image Compression Using Differential Coding and Optimization Method | |
| Kale et al. | Visually improved image compression by combining EZW encoding with texture modeling using Huffman Encoder | |
| Ramaswamy et al. | Context-based lossless image coding using EZW framework | |
| Annadurai et al. | Wavelet based color image compression using vector quantization and morphology | |
| Chen et al. | Region of interest determined by perceptual-quality and rate-distortion optimization in JPEG 2000 | |
| Yuan et al. | Context-modeled wavelet difference reduction coding based on fractional bit-plane partitioning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| ASS | Succession or assignment of patent right |
Owner name: WANG GUOQIU Free format text: FORMER OWNER: ZHONGXIN DIGITAL TECHN CO., LTD, HUNAN Effective date: 20060331 |
|
| C41 | Transfer of patent application or patent right or utility model | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20060331 Address after: 410073 Hunan province Changsha Kaifu District, Jiayuan Indus court 13 Building No. 403 Applicant after: Wang Guoqiu Address before: 410100 No. 1, No. 2, 3 District, Changsha economic and Technological Development Zone, Hunan, China Applicant before: Zhongxin Digital Techn Co., Ltd, Hunan |
|
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C41 | Transfer of patent application or patent right or utility model | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20081219 Address after: Floor 7, building MO, torch City, Changsha hi tech Industrial Development Zone, Hunan Patentee after: Hunan Qitai Information Technology Co., Ltd. Address before: Hunan province Changsha Kaifu District, Jiayuan Indus court 13 Building No. 403 Patentee before: Wang Guoqiu |
|
| ASS | Succession or assignment of patent right |
Owner name: HUNAN QITAI INFORMATION TECHNOLOGY CO., LTD. Free format text: FORMER OWNER: WANG GUOQIU Effective date: 20081219 |
|
| ASS | Succession or assignment of patent right |
Owner name: WANG GUOQIU Free format text: FORMER OWNER: HUNAN CHNTEK TECHNOLOGIES CO., LTD. Effective date: 20100908 |
|
| C41 | Transfer of patent application or patent right or utility model | ||
| COR | Change of bibliographic data |
Free format text: CORRECT: ADDRESS; FROM: 410080 7/F, BUILDING MO, HUOJUCHENG, HIGH-TECH. INDUSTRY DEVELOPMENT AREA, HUNAN PROVINCE TO: 410073 NO.13-403, WUTONGYUAN, KEDAJIAYUAN, KAIFU DISTRICT, CHANGSHA CITY, HUNAN PROVINCE |
|
| TR01 | Transfer of patent right |
Effective date of registration: 20100908 Address after: 410073 Hunan province Changsha Kaifu District, garden Indus court No. 13-403 Patentee after: Wang Guoqiu Address before: 7, building 410080, building MO, torch City, Changsha hi tech Industrial Development Zone, Hunan, China Patentee before: Hunan Qitai Information Technology Co., Ltd. |
|
| ASS | Succession or assignment of patent right |
Owner name: HUNAN HUAGUAN DIGITAL TECHNOLOGY CO., LTD. Free format text: FORMER OWNER: WANG GUOQIU Effective date: 20101110 |
|
| C41 | Transfer of patent application or patent right or utility model | ||
| COR | Change of bibliographic data |
Free format text: CORRECT: ADDRESS; FROM: 410073 NO.13-403, WUTONGYUAN, KEDAJIAYUAN, KAIFU DISTRICT, CHANGSHA CITY, HU'NAN PROVINCE TO: 410006 INSIDE OF CHANGSHA CEC SOFTWARE PARK, NO.39, JIANSHAN ROAD, YUELU DISTRICT, CHANGSHA CITY, HU'NAN PROVINCE |
|
| TR01 | Transfer of patent right |
Effective date of registration: 20101110 Address after: 410006, Changsha Electric Power Software Park, No. 39, pointed mountain road, Hunan, Changsha, Yuelu District Patentee after: Hunan Huaguan Digital Technology Co., Ltd. Address before: 410073 Hunan province Changsha Kaifu District, garden Indus court No. 13-403 Patentee before: Wang Guoqiu |
|
| ASS | Succession or assignment of patent right |
Owner name: WANG GUOQIU Free format text: FORMER OWNER: HU'NAN HUAGUAN DIGITAL TECHNOLOGY CO., LTD. Effective date: 20130109 |
|
| C41 | Transfer of patent application or patent right or utility model | ||
| COR | Change of bibliographic data |
Free format text: CORRECT: ADDRESS; FROM: 410006 CHANGSHA, HUNAN PROVINCE TO: 410073 CHANGSHA, HUNAN PROVINCE |
|
| TR01 | Transfer of patent right |
Effective date of registration: 20130109 Address after: 410073 Hunan province Changsha Kaifu District, garden Indus court 13-403 Patentee after: Wang Guoqiu Address before: 410006, Changsha Electric Power Software Park, No. 39, pointed mountain road, Hunan, Changsha, Yuelu District Patentee before: Hunan Huaguan Digital Technology Co., Ltd. |
|
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20060920 Termination date: 20160618 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |