CN108174223A - Estimate the method and system of JPEG compression bitmap quantization step - Google Patents
Estimate the method and system of JPEG compression bitmap quantization step Download PDFInfo
- Publication number
- CN108174223A CN108174223A CN201611113143.9A CN201611113143A CN108174223A CN 108174223 A CN108174223 A CN 108174223A CN 201611113143 A CN201611113143 A CN 201611113143A CN 108174223 A CN108174223 A CN 108174223A
- Authority
- CN
- China
- Prior art keywords
- coefficient sequence
- function
- quantization step
- coefficient
- quantization
- 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
- 238000013139 quantization Methods 0.000 title claims abstract description 89
- 230000006835 compression Effects 0.000 title claims abstract description 38
- 238000007906 compression Methods 0.000 title claims abstract description 38
- 238000000034 method Methods 0.000 title claims abstract description 25
- 239000011159 matrix material Substances 0.000 claims abstract description 17
- 238000006243 chemical reaction Methods 0.000 claims description 6
- 239000000284 extract Substances 0.000 claims 1
- 238000001514 detection method Methods 0.000 description 6
- 230000000694 effects Effects 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000006837 decompression Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
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
- H04N19/60—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using transform coding
- H04N19/625—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using transform coding using discrete cosine transform [DCT]
-
- 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
Abstract
The present invention relates to it is a kind of estimate JPEG compression bitmap quantization step method, including：Dct transform is carried out to the JPEG compression bitmap of input, obtains corresponding DCT coefficient matrix；To DCT coefficient matrix obtained above, DCT coefficient sequence is obtained according to different decimation in frequency；To obtained each DCT coefficient sequence, quantization error function is calculated；According to the quantization error function, function g (u are calculated to above-mentioned each DCT coefficient sequence；Cij)；It traverses all frequencies and obtains each g (u；Cij) minimum value position, to obtain the estimated value of quantization step.The invention further relates to a kind of systems for estimating JPEG compression bitmap quantization step.The present invention is without being by virtue of experience manually set threshold value, g (u；Cij) calculating C is utilizedijWhole coefficients of coefficient sequence, computation complexity are low, it is easy to accomplish.
Description
Technical field
The present invention relates to a kind of method and system for estimating JPEG compression bitmap quantization step.
Background technology
JPEG lossy compressions can effectively remove the redundancy of image, have higher compressing file rate and details
Fidelity, and the mechanism (by selecting the different compression quality factors) weighed between the two is provided, therefore in image capture
Extensive use in equipment.With the development of image-editing technology, the content of jpeg image is it is possible that being maliciously tampered and being used for illegal
Purpose, and preserved again in the form of bitmap (Bitmap) after distorting.If these bitmaps forged cannot be distinguished correctly, can
It can cause serious social danger：Public opinion may be misled as seen in news report, may be caused as court evidence
False case misjudged case.Which only distinguishes from magnanimity bitmap by distorting both inefficiency by human eye, it is also unrealistic.It is feasible at present
Solution is the technology that a kind of computer auto-detection is distorted.
Many scholars have carried out the tampering detection technology of bitmap research extensively and profoundly, it is proposed that a variety of methods.This
A little methods are usually required for knowing bitmap whole picture to be detected or even each local compression histories information, that is, need to know that bitmap is
It is no to pass through JPEG compression and if through overcompression, which kind of quantization parameter is used when estimation is compressed.JPEG compression bitmap detection side
It is fine accurate that can method and system accomplish, directly affects the performance of specific altering detecting method and system.Therefore, pass through inspection
Whether location figure (Bitmap) carries by the tampering detection that the quantization parameter of JPEG compression and estimation compress bitmap can be bitmap
For reliable compression histories information.
In entire JPEG compression and decompression process, DCT and IDCT, coding and decoding are two pairs of lossless operations, only
The loss of image information is mainly caused by quantization.In other words, quantization operation can leave quantization effect on jpeg image.Through JPEG
The bitmap of compression is there are quantization effect, and the bitmap without JPEG compression is then not present.It therefore, can by detecting quantization effect
Whether differentiate a width bitmap by JPEG compression, and it is the common side for detecting quantization effect to estimate JPEG compression bitmap quantization step
Method.
However, the method for existing estimation JPEG compression bitmap quantization step has a disadvantage that：(1) part DCT is only utilized
Coefficient and not all coefficient, therefore all available informations are not made full use of to estimate quantization step, estimation accuracy rate is not high；
(2) selection of threshold value, it is difficult to determine that an optimal threshold is all suitable for different images or even with the different frequency of piece image.
Invention content
In view of this, it is necessary to provide it is a kind of estimate JPEG compression bitmap quantization step method and system, can gram
The shortcomings that taking existing method of estimation, the tampering detection for bitmap provide foundation.
The present invention provides a kind of method for estimating JPEG compression bitmap quantization step, and this method comprises the following steps：A. it is right
The JPEG compression bitmap of input carries out dct transform, obtains corresponding DCT coefficient matrix；B. to DCT coefficient square obtained above
Battle array, DCT coefficient sequence is obtained according to different decimation in frequency；C. to obtained each DCT coefficient sequence, quantization error letter is calculated
Number；D. according to the quantization error function, function g (u are calculated to above-mentioned each DCT coefficient sequence；Cij)；E. all frequencies are traversed
Obtain each g (u；Cij) minimum value position, to obtain the estimated value of quantization step.
Wherein, the step c is specifically included：If each coefficient sequence CijN number of coefficient is shared, calculates quantization error function
It is as follows：
In above formula | | expression takes absolute value, xkIt is coefficient sequence CijK-th of element, q is the possibility value of quantization step,
Generally enable 1≤q≤M.f(q；Cij) it is with CijFor parameter, using q as the function of independent variable.
The step c is specifically included：If each coefficient sequence CijN number of coefficient is shared, it is as follows to calculate quantization error function：
In above formula | | expression takes absolute value, xkIt is coefficient sequence CijK-th of element, q is the possibility value of quantization step,
Generally enable 1≤q≤M.f(q；Cij) it is with CijFor parameter, using q as the function of independent variable.
The step c is specifically included：If each coefficient sequence CijN number of coefficient is shared, it is as follows to calculate quantization error function：
In above formula | | expression takes absolute value, xkIt is coefficient sequence CijK-th of element, q is the possibility value of quantization step,
Generally enable 1≤q≤M.f(q；Cij) it is with CijFor parameter, using q as the function of independent variable.
It is described calculating function g (u) be：g(u；Cij)=f (u；Cij)-max1≤x≤u-1f(x；Cij)。
The present invention also provides it is a kind of estimate JPEG compression bitmap quantization step system, including conversion module, abstraction module,
Computing module and estimation module, wherein：The conversion module is used to carry out dct transform to the JPEG compression bitmap of input, obtains
Corresponding DCT coefficient matrix；The abstraction module is used for DCT coefficient matrix obtained above, according to different decimation in frequency
Obtain DCT coefficient sequence；The computing module is used for each DCT coefficient sequence to obtaining, and calculates quantization error function；It is described
Computing module is additionally operable to according to the quantization error function, and function g (u are calculated to above-mentioned each DCT coefficient sequence；Cij)；It is described
Estimation module obtains each g (u for traversing all frequencies；Cij) minimum value position, to obtain the estimated value of quantization step.
Wherein, the computing module is specifically used for：If each coefficient sequence CijN number of coefficient is shared, quantization is calculated and misses
Difference function is as follows：
In above formula | | expression takes absolute value, xkIt is coefficient sequence CijK-th of element, q is the possibility value of quantization step,
Generally enable 1≤q≤M.f(q；Cij) it is with CijFor parameter, using q as the function of independent variable.
The computing module is specifically used for：If each coefficient sequence CijN number of coefficient is shared, calculates quantization error function
It is as follows：
In above formula | | expression takes absolute value, xkIt is coefficient sequence CijK-th of element, q is the possibility value of quantization step,
Generally enable 1≤q≤M.f(q；Cij) it is with CijFor parameter, using q as the function of independent variable.
The computing module is specifically used for：If each coefficient sequence CijN number of coefficient is shared, calculates quantization error function
It is as follows：
In above formula | | expression takes absolute value, xkIt is coefficient sequence CijK-th of element, q is the possibility value of quantization step,
Generally enable 1≤q≤M.f(q；Cij) it is with CijFor parameter, using q as the function of independent variable.
It is described calculating function g (u) be：g(u；Cij)=f (u；Cij)-max1≤x≤u-1f(x；Cij)。
The present invention can overcome the shortcomings that existing method of estimation, and the tampering detection for bitmap provides foundation.The present invention's has
Beneficial effect includes：(1) it without threshold value is by virtue of experience manually set, avoids introducing inappropriate parameter setting.(2) whole is utilized
DCT coefficient carries out quantization step estimation.(3) it is not related to difficult mathematical operation, computation complexity is low, it is easy to accomplish.
Description of the drawings
Fig. 1 is the flow chart of the method for present invention estimation JPEG compression bitmap quantization step；
Fig. 2 is function g (u of the present invention；Cij) an exemplary schematic diagram.
Fig. 3 is the hardware architecture diagram of the system of present invention estimation JPEG compression bitmap quantization step.
Specific embodiment
Below in conjunction with the accompanying drawings and specific embodiment the present invention is described in further detail.
As shown in fig.1, it is the job stream of the method preferred embodiment of present invention estimation JPEG compression bitmap quantization step
Cheng Tu.
Step S1 carries out dct transform to the JPEG compression bitmap of input, obtains corresponding DCT coefficient matrix.It is specific and
Speech：
To the JPEG compression bitmap of input, the dct transform of 8x8 piecemeals is carried out, obtains corresponding DCT coefficient matrix.
To DCT coefficient matrix obtained above, DCT coefficient sequence is obtained according to different decimation in frequency by step S2.Specifically
Including：
Due to a shared 8x8=64 frequencies, all coefficients of same frequency are extracted and form a coefficient sequence,
Therefore 64 coefficient sequences can be obtained altogether, is denoted as Cij, 0≤i, j≤7.
Step S3 to obtained each DCT coefficient sequence, calculates quantization error function.Specifically：
If each coefficient sequence CijN number of coefficient is shared, it is as follows to calculate quantization error function：
In above formula | | expression takes absolute value, xkIt is coefficient sequence CijK-th of element, q is the possibility value of quantization step,
Generally enable 1≤q≤M.f(q；Cij) it is with CijFor parameter, using q as the function of independent variable.
In other embodiments of the invention, the quantization error function can also be：
Or
According to the quantization error function, function g (u are calculated to above-mentioned each DCT coefficient sequence by step S4；Cij).Specifically
For：
It is as follows that the present embodiment calculates function g (u)：
If u=1, g (u are enabled；Cij)=f (u；Cij).The physical meaning of above formula：By f (u；Cij) subtract all values before it
Maximum value.f(u；Cij) general trend be to become larger with the increase of u, but can be in quantization step and its multiple or approximate number
Place takes local minimum.Therefore, function g (u；Cij) minimum value can occur in the position of quantization step, by detecting this
A minimum value, it is possible to obtain the estimated value of quantization step.
g(u；Cij) an example as shown in Figure 2.g(u；Cij) curve at frequency (i, j)=(2,2) is as schemed
Show.When curve is minimized, corresponding x coordinate is exactly the quantization step to be estimated of the frequency for 12, i.e., 12.
Step S5 traverses all frequencies and obtains each g (u；Cij) minimum value position, as quantization step estimated value.
Specifically：
The present embodiment traverses all 0≤i of frequency, and j≤7 calculate g (u；Cij), g (u are found out respectively；Cij) in different frequency
The corresponding position of lower minimum value is to get to JPEG compression bitmap quantization step estimated value qij.The quantization step estimated value qijOne
Totally 64, be the matrix of 8x8.
As shown in fig.3, it is the hardware architecture diagram of the system 10 of present invention estimation JPEG compression bitmap quantization step.This is
System includes conversion module 101, abstraction module 102, computing module 103 and estimation module 104.
The conversion module 101 is used to carry out dct transform to the JPEG compression bitmap of input, obtains corresponding DCT coefficient
Matrix.Specifically：
To the JPEG compression bitmap of input, the dct transform of 8x8 piecemeals is carried out, obtains corresponding DCT coefficient matrix.
The abstraction module 102 is used for DCT coefficient matrix obtained above, and DCT is obtained according to different decimation in frequency
Coefficient sequence.It specifically includes：
Due to a shared 8x8=64 frequencies, all coefficients of same frequency are extracted and form a coefficient sequence,
Therefore 64 coefficient sequences can be obtained altogether, is denoted as Cij, 0≤i, j≤7.
The computing module 103 is used for each DCT coefficient sequence to obtaining, and calculates quantization error function.Specifically：
If each coefficient sequence CijN number of coefficient is shared, it is as follows to calculate quantization error function：
In above formula | | expression takes absolute value, xkIt is coefficient sequence CijK-th of element, q is the possibility value of quantization step,
Generally enable 1≤q≤M.f(q；Cij) it is with CijFor parameter, using q as the function of independent variable.
In other embodiments of the invention, the quantization error function can also be：
Or
The computing module 103 is additionally operable to according to the quantization error function, and letter is calculated to above-mentioned each DCT coefficient sequence
Number g (u；Cij).Specifically：
It is as follows that the present embodiment calculates function g (u)：
If u=1, g (u are enabled；Cij)=f (u；Cij).The physical meaning of above formula：By f (u；Cij) subtract all values before it
Maximum value.f(u；Cij) general trend be to become larger with the increase of u, but can be in quantization step and its multiple or approximate number
Place takes local minimum.Therefore, function g (u；Cij) minimum value can occur in the position of quantization step, by detecting this
A minimum value, it is possible to obtain the estimated value of quantization step.g(u；Cij) an example as shown in Figure 2.
The estimation module 104 obtains each g (u for traversing all frequencies；Cij) minimum value position, as quantization step
Long estimated value.Specifically：
The present embodiment traverses all 0≤i of frequency, and j≤7 calculate g (u；Cij), g (u are found out respectively；Cij) in different frequency
The corresponding position of lower minimum value is to get to JPEG compression bitmap quantization step estimated value qij.The quantization step estimated value qijOne
Totally 64, be the matrix of 8x8.
Function g (the u of the present invention；Cij) f (u are utilized；Cij) general trend information, directly by finding g (u；Cij)
Minimum value can estimate quantization step, without artificially given threshold by rule of thumb.Meanwhile g (u；Cij) calculating C is utilizedij
Whole coefficients of coefficient sequence, in terms of information utilization angle, the present invention is more abundant to the utilization of coefficient.
Although the present invention is described with reference to current better embodiment, those skilled in the art should be able to manage
Solution, above-mentioned better embodiment is only used for illustrating the present invention, is not used for limiting protection scope of the present invention, any in the present invention
Spirit and spirit within, any modification for being done, equivalence replacement, improvements etc. should be included in the right guarantor of the present invention
Within the scope of shield.
Claims (10)
- A kind of 1. method for estimating JPEG compression bitmap quantization step, which is characterized in that the method comprising the steps of：A. dct transform is carried out to the JPEG compression bitmap of input, obtains corresponding DCT coefficient matrix；B. to DCT coefficient matrix obtained above, DCT coefficient sequence is obtained according to different decimation in frequency；C. to obtained each DCT coefficient sequence, quantization error function is calculated；D. according to the quantization error function, function g (u are calculated to above-mentioned each DCT coefficient sequence；Cij)；E. all frequencies are traversed and obtain each g (u；Cij) minimum value position, to obtain the estimated value of quantization step.
- 2. the method as described in claim 1, which is characterized in that the step c is specifically included：If each coefficient sequence CijN number of coefficient is shared, it is as follows to calculate quantization error function：In above formula | | expression takes absolute value, xkIt is coefficient sequence CijK-th of element, q is the possibility value of quantization step, generally Enable 1≤q≤M.f(q；Cij) it is with CijFor parameter, using q as the function of independent variable.
- 3. the method as described in claim 1, which is characterized in that the step c is specifically included：If each coefficient sequence CijN number of coefficient is shared, it is as follows to calculate quantization error function：In above formula | | expression takes absolute value, xkIt is coefficient sequence CijK-th of element, q is the possibility value of quantization step, generally Enable 1≤q≤M.f(q；Cij) it is with CijFor parameter, using q as the function of independent variable.
- 4. the method as described in claim 1, which is characterized in that the step c is specifically included：If each coefficient sequence CijN number of coefficient is shared, it is as follows to calculate quantization error function：In above formula | | expression takes absolute value, xkIt is coefficient sequence CijK-th of element, q is the possibility value of quantization step, generally Enable 1≤q≤M.f(q；Cij) it is with C1jFor parameter, using q as the function of independent variable.
- 5. the method as described in Claims 2 or 3 or 4, which is characterized in that it is described calculating function g (u) be：g(u；Cij)=f (u；Cij)-max1≤x≤u-1f(x；Cij)。
- 6. a kind of system for estimating JPEG compression bitmap quantization step, which is characterized in that the system includes conversion module, extracts mould Block, computing module and estimation module, wherein：The conversion module is used to carry out dct transform to the JPEG compression bitmap of input, obtains corresponding DCT coefficient matrix；The abstraction module is used for DCT coefficient matrix obtained above, and DCT coefficient sequence is obtained according to different decimation in frequency Row；The computing module is used for each DCT coefficient sequence to obtaining, and calculates quantization error function；The computing module is additionally operable to according to the quantization error function, and function g (u are calculated to above-mentioned each DCT coefficient sequence； Cij)；The estimation module obtains each g (u for traversing all frequencies；Cij) minimum value position, to obtain quantization step Estimated value.
- 7. system as claimed in claim 6, which is characterized in that the computing module is specifically used for：If each coefficient sequence CijN number of coefficient is shared, it is as follows to calculate quantization error function：In above formula | | expression takes absolute value, xkIt is coefficient sequence CijK-th of element, q is the possibility value of quantization step, generally Enable 1≤q≤M.f(q；Cij) it is with CijFor parameter, using q as the function of independent variable.
- 8. system as claimed in claim 6, which is characterized in that the computing module is specifically used for：If each coefficient sequence CijN number of coefficient is shared, it is as follows to calculate quantization error function：In above formula | | expression takes absolute value, xkIt is coefficient sequence CijK-th of element, q is the possibility value of quantization step, generally Enable 1≤q≤M.f(q；Cij) it is with CijFor parameter, using q as the function of independent variable.
- 9. system as claimed in claim 6, which is characterized in that the computing module is specifically used for：If each coefficient sequence CijN number of coefficient is shared, it is as follows to calculate quantization error function：In above formula | | expression takes absolute value, xkIt is coefficient sequence CijK-th of element, q is the possibility value of quantization step, generally Enable 1≤q≤M.f(q；Cij) it is with CijFor parameter, using q as the function of independent variable.
- 10. the method as described in claim 7 or 8 or 9, which is characterized in that it is described calculating function g (u) be：g(u；Cij)=f (u；Cij)-max1≤x≤u-1f(x；Cij)。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611113143.9A CN108174223B (en) | 2016-12-07 | 2016-12-07 | Method and system for estimating quantization step size of JPEG (joint photographic experts group) compressed bitmap |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611113143.9A CN108174223B (en) | 2016-12-07 | 2016-12-07 | Method and system for estimating quantization step size of JPEG (joint photographic experts group) compressed bitmap |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108174223A true CN108174223A (en) | 2018-06-15 |
| CN108174223B CN108174223B (en) | 2020-07-24 |
Family
ID=62526092
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611113143.9A Active CN108174223B (en) | 2016-12-07 | 2016-12-07 | Method and system for estimating quantization step size of JPEG (joint photographic experts group) compressed bitmap |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108174223B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070195106A1 (en) * | 2006-02-17 | 2007-08-23 | Microsoft Corporation | Detecting Doctored JPEG Images |
| CN102226920A (en) * | 2011-06-03 | 2011-10-26 | 贵州大学 | Cutting-resistant JPEG image compression history and synthetic tamper detection method |
| CN102801975A (en) * | 2012-07-31 | 2012-11-28 | 李斌 | Method and device for estimating processing of quantifying step size of pictures |
| CN105120294A (en) * | 2015-06-26 | 2015-12-02 | 中国电子科技集团公司第二十八研究所 | JPEG format image source identification method |
-
2016
- 2016-12-07 CN CN201611113143.9A patent/CN108174223B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070195106A1 (en) * | 2006-02-17 | 2007-08-23 | Microsoft Corporation | Detecting Doctored JPEG Images |
| CN102226920A (en) * | 2011-06-03 | 2011-10-26 | 贵州大学 | Cutting-resistant JPEG image compression history and synthetic tamper detection method |
| CN102801975A (en) * | 2012-07-31 | 2012-11-28 | 李斌 | Method and device for estimating processing of quantifying step size of pictures |
| CN105120294A (en) * | 2015-06-26 | 2015-12-02 | 中国电子科技集团公司第二十八研究所 | JPEG format image source identification method |
Non-Patent Citations (2)
| Title |
|---|
| ZHIGANG FAN，ET AL.: "Identification of Bitmap Compression History:", 《IEEE TRANSACTIONS ON IMAGE PROCESSING》 * |
| 周响金 等: "基于JPEG量化失真的合成图像盲检测", 《计算机应用研究》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108174223B (en) | 2020-07-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP2786342B1 (en) | Texture masking for video quality measurement | |
| US10034005B2 (en) | Banding prediction for video encoding | |
| US20200327701A1 (en) | Image encoding method and apparatus and image decoding method and apparatus | |
| US8406462B2 (en) | Signature derivation for images | |
| CN102413328B (en) | Double compression detection method and system of joint photographic experts group (JPEG) image | |
| WO2016082277A1 (en) | Video authentication method and apparatus | |
| Wang et al. | Novel spatio-temporal structural information based video quality metric | |
| CN113810611B (en) | Data simulation method and device for event camera | |
| WO2014070273A1 (en) | Recursive conditional means image denoising | |
| CN103067713B (en) | Method and system of bitmap joint photographic experts group (JPEG) compression detection | |
| US9268791B2 (en) | Method and apparatus for image processing and computer readable medium | |
| CN105227964B (en) | Video-frequency identifying method and system | |
| WO2019024160A1 (en) | Image tampering detection method and system, electronic device, and readable storage medium | |
| Li et al. | Quantization step estimation for JPEG image forensics | |
| Patil | A survey on image quality assessment techniques, challenges and databases | |
| Fu et al. | Screen content image quality assessment using Euclidean distance | |
| WO2019087383A1 (en) | Crowd density calculation device, crowd density calculation method and crowd density calculation program | |
| CN109544557A (en) | Principal component analysis conversion method and device based on block | |
| CN108230411B (en) | Method and device for detecting tampered image | |
| CN108174223A (en) | Estimate the method and system of JPEG compression bitmap quantization step | |
| CN102742271B (en) | The compression of images using discrete cosine transform detects the artifact of quantizing noise | |
| CN116797510A (en) | Image processing method, device, computer equipment and storage medium | |
| Yang et al. | Transparent composite model for large scale image/video processing | |
| CN107818325A (en) | Image sparse method for expressing based on integrated dictionary learning | |
| CN104796726A (en) | Method for detecting image compression traces |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20180615 Assignee: Shenzhen three pen Technology Co.,Ltd. Assignor: SHENZHEN INSTITUTES OF ADVANCED TECHNOLOGY CHINESE ACADEMY OF SCIENCES Contract record no.: X2023980043748 Denomination of invention: Method and System for Estimating the Quantization Step of JPEG Compressed Bitmaps Granted publication date: 20200724 License type: Common License Record date: 20231019 |
|
| EE01 | Entry into force of recordation of patent licensing contract |