CN106327470B - A kind of detection method and device of image enhancement processing - Google Patents
A kind of detection method and device of image enhancement processing Download PDFInfo
- Publication number
- CN106327470B CN106327470B CN201510368803.7A CN201510368803A CN106327470B CN 106327470 B CN106327470 B CN 106327470B CN 201510368803 A CN201510368803 A CN 201510368803A CN 106327470 B CN106327470 B CN 106327470B
- Authority
- CN
- China
- Prior art keywords
- enhancing
- image
- estimates
- dct coefficient
- parameters
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20048—Transform domain processing
- G06T2207/20052—Discrete cosine transform [DCT]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30168—Image quality inspection
Abstract
The present invention is suitable for technical field of image processing, provides a kind of detection method and device of image enhancement processing, comprising: calculate the quantized-dct DCT coefficient distribution period of described image；According to the quantization DCT coefficient distribution period, the enhancing estimates of parameters of described image is calculated；Judge whether the enhancing estimates of parameters is equal to 1；If the enhancing estimates of parameters is equal to 1, described image is detected as enhancing the image of processing without compression domain；If the enhancing estimates of parameters is not equal to 1, described image is detected as enhancing the image of processing by compression domain.The present invention is for given image, the enhancing parameter of its compression domain is estimated, if the enhancing parameter estimated is not equal to 1, then think enhancing processing of the image Jing Guo compression domain, and the image before enhancing can be further restored according to the estimated value of the enhancing parameter, to complete to enhance processing detection to the compression domain of image with detection accuracy with higher estimation.
Description
Technical field
The invention belongs to technical field of image processing more particularly to a kind of detection method and device of image enhancement processing.
Background technique
As one of common image processing techniques, image enhancement technique passes through the contrast for increasing image border texture,
Achieve the purpose that highlight image detail, image enhancement technique is broadly divided into pixel domain enhancing and compression domain enhances two kinds.
Piece image may be saved locally, and share the Internet user into internet after enhancing processing
After downloading image, it may be desirable to recognize whether the image that it is downloaded have passed through enhancing processing, thereby increases and it is possible to it is desirable that from process
Original image before restoring enhancing processing in the image of enhancing processing, to obtain the pictorial information of most original, currently, being directed to
Whether the image that enhancing processing has been carried out using Image in Compressed Domain enhancing technology, be still capable of detecting when it without effective method
It is handled by enhancing.
Summary of the invention
The embodiment of the present invention provides a kind of detection method and device of image enhancement processing, it is intended to propose that one kind is able to detect
Whether image have passed through the technical solution of compression domain enhancing processing out.
The embodiments of the present invention are implemented as follows, a kind of detection method of image enhancement processing, comprising:
Calculate the quantized-dct DCT coefficient distribution period of described image；
According to the quantization DCT coefficient distribution period, the enhancing estimates of parameters of described image is calculated；
Judge whether the enhancing estimates of parameters is equal to 1；
If the enhancing estimates of parameters is equal to 1, described image is detected as enhancing the figure of processing without compression domain
Picture；If the enhancing estimates of parameters is not equal to 1, described image is detected as enhancing the image of processing by compression domain.
The another object of the embodiment of the present invention is to provide a kind of detection device of image enhancement processing, comprising:
First computing unit, for calculating the quantized-dct DCT coefficient distribution period of described image；
Second computing unit, for according to the quantization DCT coefficient distribution period, the enhancing parameter for calculating described image to be estimated
Evaluation；
Judging unit, for judging whether the enhancing estimates of parameters is equal to 1；
Testing result output unit, if for the enhancing estimates of parameters equal to 1, described image be detected as without
The image of overcompression domain enhancing processing；If the enhancing estimates of parameters is not equal to 1, described image is detected as through overcompression
The image of domain enhancing processing.
In the embodiment of the present invention, for given image, the enhancing parameter of its compression domain is estimated, if estimate
Enhance parameter and be not equal to 1, then it is assumed that enhancing of the image Jing Guo compression domain is handled, and can according to the estimated value of the enhancing parameter into
Image before restoring to one step enhancing, to be completed at the compression domain enhancing to image with higher estimation with detection accuracy
Reason detection.
Detailed description of the invention
It to describe the technical solutions in the embodiments of the present invention more clearly, below will be to embodiment or description of the prior art
Needed in attached drawing be briefly described, it should be apparent that, the accompanying drawings in the following description is only of the invention some
Embodiment for those of ordinary skill in the art without any creative labor, can also be according to these
Attached drawing obtains other attached drawings.
Fig. 1 is that the frequency band for 8 × 8 nonoverlapping DCT coefficient matrixes that the prior art provides divides schematic diagram；
Fig. 2 is the implementation flow chart of the detection method of image enhancement processing provided in an embodiment of the present invention；
Fig. 3 is the specific implementation of the implementation process S201 of the detection method of image enhancement processing provided in an embodiment of the present invention
Flow chart；
Fig. 4 is the specific implementation of the implementation process S202 of the detection method of image enhancement processing provided in an embodiment of the present invention
Flow chart；
Fig. 5 be another embodiment of the present invention provides image enhancement processing detection method implementation flow chart；
Fig. 6 is original image provided in an embodiment of the present invention, enhances image and the corresponding intuitive comparison for going back original image
Effect picture；
Fig. 7 is the structural block diagram of the detection device of image enhancement processing provided in an embodiment of the present invention.
Specific embodiment
In being described below, for illustration and not for limitation, the tool of such as particular system structure, technology etc is proposed
Body details understands the embodiment of the present invention to cut thoroughly.However, it will be clear to one skilled in the art that there is no these specific
The present invention also may be implemented in the other embodiments of details.In other situations, it omits to well-known system, device, electricity
The detailed description of road and method, in case unnecessary details interferes description of the invention.
In the embodiment of the present invention, for given image, the enhancing parameter of its compression domain is estimated, if estimate
Enhance parameter and be not equal to 1, then it is assumed that enhancing of the image Jing Guo compression domain is handled, and can according to the estimated value of the enhancing parameter into
Image before restoring to one step enhancing, to be completed at the compression domain enhancing to image with higher estimation with detection accuracy
Reason detection.
It should be noted that the image for carrying out enhancing processing detection is a width bitmap (Bitmap) in the embodiment of the present invention
Image, the image may have following three kinds of possibility: (1) without JPEG compression；(2) it have passed through JPEG compression but without overcompression
Domain enhancing；(3) have passed through JPEG compression simultaneously and have passed through compression domain enhancing, the embodiment of the present invention can to bitmap images whether
It has passed through JPEG compression simultaneously and compression domain enhancing detected.Wherein, JPEG, i.e. joint photographic experts group (Joint
Photographic Experts Group).JPEG is current most widely used Standard of image compression.The compression standard is detailed
It defines the specific steps of compression and decompression: in compression, image pixel matrix being divided into the nonoverlapping more of 8x8 first
A piecemeal, and discrete cosine transform (Discrete Cosine Transform, DCT) is carried out to each piecemeal, obtain DCT system
Number；Then, according to quantization table (for example, 8 × 8 quantization step matrix) set by user, DCT coefficient block is quantified；Most
Afterwards, entropy coding is carried out to resulting quantization DCT coefficient, and is write code stream as jpeg file；Decompression is the inverse process of compression, first
Entropy decoding first is carried out to the code stream in jpeg image file, then to quantization DCT coefficient inverse quantization, finally, to inverse quantization DCT system
Number carries out inverse discrete cosine transformation (IDCT), and restores picture element matrix after carrying out floor operation.It should be noted that if to
Compression image is color image, then need before compression with color space conversion and sub-sampling are carried out after decompression.The present invention is implemented
Scheme elaboration is carried out just for gray level image in example.
In order to better understand jpeg image provided by this programme compression domain enhancing processing detection and reduction principle,
First the image enhancement principle of jpeg image compression domain is summarized as follows below:
In paper " Image enhancement using a contrast measure in the compressed
In domain, IEEE Signal Processing Letters, vol.10, no.10, pp.289-292.2003 ", Tang etc.
People propose compression domain comparison estimates, be defined as the average amplitude of high frequency band and the average amplitude of low-frequency band and the ratio between,
And the compression domain enhancing processing for having carried out jpeg image is estimated based on the comparison:
In the compression process of jpeg image, 8 × 8 nonoverlapping DCT coefficient matrixes, these DCT are obtained after DCT processing
Coefficient matrix obtains quantization DCT coefficient after being quantized, and uses dI, jIndicate that frequency is the quantization DCT system of (i, j) in DCT coefficient matrix
8 × 8 nonoverlapping DCT coefficient matrixes are divided into 15 frequency bands by back-diagonal direction by number, 0≤i, j≤7, Tang et al.,
N-th of frequency band is expressed as Bn={ dI, j| i+j=n }, wherein 0≤n≤14, as shown in Figure 1, having irised out frequency band B in figure2And B9, n
Bigger, frequency locating for frequency band is higher.
Tang et al. is as follows in the enhancing processing step of compression domain to jpeg image:
1, C is estimated in definition comparisonn:
Wherein,CnIt is meant that n-th of frequency band
The ratio of absolute coefficient mean value and the sum of the absolute coefficient mean value of preceding n-1 frequency band, CnShow the absolute coefficient of n-th of frequency band
Opposite value size.
2, by improving the accounting of jpeg image radio-frequency component, it can achieve purpose (the image high frequency of enhancing image detail
Ingredient corresponds to image detail), Tang et al. enables modified comparison estimateIt is that C is estimated in the preceding comparison of modificationnλ (λ be enhancing
Parameter) times:
And then it can derive:
Enhanced quantization DCT coefficient is indicated in formula, is quantization before enhancing
The λ H of DCT coefficientnTimes.In other words, carrying out the JPEG compression domain that parameter is λ to image enhances, and is exactly to former quantization DCT coefficient
dI, jMake λ HnMultiply operation again, wherein HnIt is determined by the statistical property of image block itself.In addition, the overall situation in order to avoid image is bright
Degree shifts, and Tang et al. keeps DC coefficient d0,0It is constant, i.e.,Therefore have
Based on the above-mentioned principle for carrying out compression domain enhancing processing to jpeg image, it can derive and jpeg image is pressed
The influence that the enhancing processing of contracting domain is distributed the image quantization DCT coefficient:
In jpeg image compression process, non-quantized DCT coefficientIt is q by step-lengthI, j(qI, jIt is set from user
Quantization matrix) quantization operation after obtain quantization DCT coefficient dI, j.This quantizing process may be expressed as:For positive integer, [] indicates that arest neighbors is rounded, that is, rounds up, in formula,It is one
Integer, therefore dI, jOnly take qI, jIntegral multiple, we claim dI, jDistribution period, i.e., quantization DCT coefficient distribution period be qI, j。
Compression domain according to jpeg image described above enhances principle, compression domain enhancing treated quantization DCT coefficientIt can be expressed as follows:
Similar,It is an integer, therefore, enhanced quantization DCT coefficientDistribution period be TI, j=λ
HnqI, j.In other words, the distribution period of quantization DCT coefficient is q before enhancingI, j, and the distribution period of quantization DCT coefficient is after enhancing
λHnqI, j, and it is related with enhancing parameter lambda.
It can discuss in two kinds of situation to the distribution period of n-th of frequency band.
(1) as n=i+j=1, due toH1=1, then have
TI, j=λ H1qI, j=λ qI, j；
(2) as n=i+j >=2, if meeting,Then have
TI, j=λ HnqI, j≈λqI, j,
Due to E0And EtIt can not observe, therefore when carrying out condition judgement, only judgeWhether meet.
By above-mentioned derivation it is recognised that the enhancing for carrying out compression domain to jpeg image is handled, quantization DCT coefficient can be changed
The periodicity of distribution: if jpeg image is handled without the enhancing of compression domain, quantization DCT coefficient distribution period is qI, j；If
Jpeg image have passed through the enhancing processing of compression domain, then following change will occur for its quantization DCT coefficient distribution period:
(1) for frequency band B1, the distribution period of quantization DCT coefficient becomes TI, j=λ qI, j, that is, become original λ times, λ is
Enhance parameter, qI, jThe quantization DCT coefficient distribution period before enhancing processing is carried out for the jpeg image；(2) for Bn(n >=
2), if meetingWhen this specified conditions (> > expression is much larger than), also there is TI, j≈λqI, j, therefore, according to
The periodicity of quantization DCT coefficient distribution, it is estimated that the enhancing parameter of jpeg image, and according to enhancing parameter whether be 1 come
Infer whether the jpeg image have passed through the enhancing processing of compression domain.
Illustrate the detection method of image enhancement processing provided by the present invention below by specific embodiment:
Fig. 2 shows the implementation processes of the detection method of image enhancement processing provided in an embodiment of the present invention, and details are as follows:
In S201, calculate jpeg image quantized-dct (Discrete Cosine Transform,
DCT) coefficient distribution period.
Specific calculation method is as shown in Figure 3:
In S301, the picture element matrix of described image is divided into 8 × 8 blocks of pixels not overlapped.
In S302, DCT processing is carried out to each blocks of pixels, obtains the 8 × 8 of each blocks of pixels
DCT coefficient matrix.
In S303, frequency band B is extractednQuantization DCT coefficient sequenceN=i+j, 0≤n≤14.
Specifically, to frequency band Bn(n=i+j=1) coefficient for extracting each frequency (i, j) constitutes 2 sequences
To frequency band Bn(1 < i+j≤W, 2≤W≤14) first reject and are unsatisfactory for conditionCoefficient block, from remaining system
In several piece, the coefficient of identical frequency is extracted, constitutes coefficient sequenceNote forms w coefficient sequence altogether.For narration
It is convenient, it is total to form 2+w coefficient sequence for 1≤i+j≤W after all sequences extract
In S304, the quantization DCT coefficient sequence is calculated separatelyGoal-selling function E (xl) functional value,
By finding the goal-selling function E (xl) position that obtains Local Minimum functional value, determine quantization DCT coefficient distribution
Period.
Goal-selling function E (the xl) it is defined as follows:
In above formula, m is quantization DCT coefficient sequenceThe number of middle coefficient, k-th of coefficient is expressed as in sequence
K=1,2...m.xlIn candidate section { x1, x2... xLIn traverse value, can be manually set.{x1, x2... xLValue between
Every smaller, quantization DCT coefficient sequenceDistribution period estimation it is more accurate.Above-mentioned goal-selling function E (xl) base
Present principles are: xlIn candidate section { x1, x2... xLIn traverse value, work as xlWhen taking the approximate number of distribution period, objective function E
(xl) local minimum will be obtained, therefore by finding E (xl) obtain local minimum position, can be inferred that quantization DCT system
Number SequenceDistribution period.
Preferably, in order to simplify calculating and improve the estimation accuracy in quantization DCT coefficient period, before S304, to pre-
If objective function E (xl) deformation are as follows:
F(xl)=E (xl)-max{E(x1), E (x2) ... E (xl-1),
The quantization DCT coefficient sequence is calculated separately againF (xl) functional value, finally, according to F (xl) can count altogether
Calculation obtains 2+w quantization DCT coefficient distribution period1≤i+j≤W。
In goal-selling function E (xl) and its deformed F (xl) in, the l is an integer index value, the value model of l
Precision needed for depending on calculating quantization DCT coefficient distribution period is enclosed, for example, if the quantization DCT coefficient to be calculated distribution week
The range of phase is preset as { 0.1,0.11,0.12 ... 39.99,40 }, due to sharing 3991 different values in this range,
Therefore the value range of l is { 1,2 ... 3991 }, wherein x1=0.1, x2=0.11 ..., x3991=40.
In S202, according to the quantization DCT coefficient distribution period, the enhancing estimates of parameters of described image is calculated.
Specific enhancing method for parameter estimation is as shown in Figure 4:
In S401, the functional value of goal-selling function G (x) is calculated according to the quantization DCT coefficient distribution period.
Wherein, the functional form of the objective function G (x) includes but is not limited to three kinds following:
1、
2、
3、
In objective function G (x), variable x is divided into S between the discrete value of α≤x≤β, value, wherein the value of α, β and S can
It presets.
Due to when x levels off to the enhancing parameter lambda to be estimated, G (x) will level off to 0, therefore, as x=λ, the above-mentioned 1st
Global minimum will be obtained with the G (x) of the 3rd kind of form, the G (x) of above-mentioned 2nd kind of form will obtain global maximum.Based on upper
Principle is stated, the estimated value of the enhancing parameter of image can be calculated
Corresponding to the functional form of above-mentioned three kinds of objective function G (x), the corresponding estimated value for enhancing parameterIt is respectively as follows:
1、
2、
3、
In S402, the value of x when by the functional value minimum of the goal-selling function G (x) is retrieved as described image
Enhance estimates of parameters.
In S203, judge whether the enhancing estimates of parameters is equal to 1.
In S204, if the enhancing estimates of parameters is equal to 1, described image is detected as enhancing without compression domain
The image of processing；If the enhancing estimates of parameters is not equal to 1, described image is detected as enhancing processing by compression domain
Image.
It further, can also be to the jpeg image after detecting that jpeg image have passed through compression domain enhancing processing
It is restored, as shown in figure 5, if the enhancing estimates of parameters is not equal to 1, the method also includes:
S205 carries out compression domain enhancing processing to described image using the reciprocal of enhancing estimates of parameters, restores
Described image is by the original image before compression domain enhancing processing.
It is as follows to have passed through the principle that the image that compression domain enhancing is handled is restored in S204:
If piece image is handled by the JPEG compression domain enhancing that enhancing parameter is λ, only need to be to diagram picture again
Carrying out the JPEG compression domain that primary enhancing parameter is 1/ λ enhances, i.e., resilient to obtain original image.Remember the quantization of non-reinforced image
DCT coefficient is dI, j, after the enhancing processing that enhanced parameter is λ, after quantization DCT coefficient becomes the enhancing using 1/ λ,
Quantization DCT coefficient becomes will be proven below
Enhance algorithm according to the JPEG compression domain that Tang et al. is proposed, satisfaction is estimated in the comparison of each image:
Have again simultaneously,
In conjunction with above-mentioned two formula, coefficientIt is represented by
Card is finished.
Next, carrying out Performance Evaluation to technical solution provided in an embodiment of the present invention by two groups of experiments.
Firstly, enhancing the estimation accuracy of parameter by first group of experimental evaluation:
It should be noted that since the estimation accuracy of enhancing parameter also reflects the standard of image enhancement processing detection simultaneously
True property (whether detection piece image passes through the enhancing of JPEG compression domain), therefore no longer additionally provide the standard of image enhancement processing detection
The detection data of true property.
The enhancing parameter lambda of test is being gathered:
{ 0.60,0.70,0.80,0.90,1.00,1.10,1.20,1.30,1.40,1.50 } traverse value, to having a size of
512 × 512, the standard picture that compression quality is 70 and 90 carries out the image enhancement in JPEG compression domain.Standard picture includes:
Baboon, Boat, F15, House, Lena, Peppers, Telephone.
The parameter that the present invention program and system is arranged is as follows:
(1) S303, the W=3 in the calculating of quantization DCT coefficient distribution period, therefore w=7.
(2) S304 in the calculating of quantization DCT coefficient distribution period, candidate section { x1, x2... xLBe 1.00,
1.01,1.02..., 20.00 }.
(3) enhance the S401, α=0.5, β=2.0, S=0.01 in the estimation of parameter.
Table 1 is shown as QF=90, the practical comparison result for enhancing parameter and estimating parameter:
Table 1
| Actual parameter | 0.60 | 0.70 | 0.80 | 0.90 | 1.00 | 1.10 | 1.20 | 1.30 | 1.40 | 1.50 |
| Baboon | 0.59 | 0.69 | 0.80 | 0.90 | 1.00 | 1.10 | 1.20 | 1.31 | 1.41 | 1.51 |
| Boat | 0.59 | 0.69 | 0.80 | 0.90 | 1.00 | 1.10 | 1.20 | 1.31 | 1.41 | 1.51 |
| F15 | 0.59 | 0.69 | 0.80 | 0.90 | 1.00 | 1.10 | 1.20 | 1.31 | 1.41 | 1.51 |
| House | 0.59 | 0.69 | 0.80 | 0.90 | 1.00 | 1.10 | 1.20 | 1.31 | 1.41 | 1.51 |
| Lena | 0.59 | 0.69 | 0.80 | 0.90 | 1.00 | 1.10 | 1.20 | 1.31 | 1.41 | 1.51 |
| Peppers | 0.59 | 0.69 | 0.80 | 0.90 | 1.00 | 1.10 | 1.20 | 1.31 | 1.41 | 1.51 |
| Telephone | 0.59 | 0.69 | 0.80 | 0.90 | 1.00 | 1.10 | 1.20 | 1.31 | 1.41 | 1.51 |
| Mean error | 1.00% | 1.00% | 0 | 0 | 0 | 0 | 0 | 1.00% | 1.00% | 1.00% |
Table 2 is shown as QF=70, the practical comparison result for enhancing parameter and estimating parameter:
Table 2
| Actual parameter | 0.60 | 0.70 | 0.80 | 0.90 | 1.00 | 1.10 | 1.20 | 1.30 | 1.40 | 1.50 |
| Baboon | 0.59 | 0.69 | 0.80 | 0.90 | 1.00 | 1.10 | 1.20 | 1.31 | 1.41 | 1.51 |
| Boat | 0.59 | 0.69 | 0.80 | 0.90 | 1.00 | 1.10 | 1.20 | 1.31 | 1.41 | 1.51 |
| F15 | 0.59 | 0.70 | 0.80 | 0.90 | 1.00 | 1.10 | 1.20 | 1.30 | 1.41 | 1.51 |
| House | 0.59 | 0.69 | 0.80 | 0.90 | 1.00 | 1.10 | 1.20 | 1.30 | 1.41 | 1.51 |
| Lena | 0.59 | 0.69 | 0.80 | 0.90 | 1.00 | 1.10 | 1.20 | 1.31 | 1.41 | 1.51 |
| Peppers | 0.59 | 0.69 | 0.80 | 0.90 | 1.00 | 1.10 | 1.20 | 1.30 | 1.41 | 1.51 |
| Telephone | 0.59 | 0.69 | 0.80 | 0.90 | 1.00 | 1.10 | 1.20 | 1.30 | 1.41 | 1.51 |
| Mean error | 1.00% | 0.86% | 0 | 0 | 0 | 0 | 0 | 0.43% | 1.00% | 1.00% |
The experimental results showed that i.e. λ is { 0.80,0.90,1.00,1.10,1.20 } when compression domain enhancing parameter is little
When, the estimated value for enhancing parameter is equal to actual enhancing parameter, and estimation is accurate；If enhancing, parameter is larger, and this programme is to enhancing
The estimation mean error of parameter is no more than 1%.It is very high accurate that experiment shows that this programme has the estimation of enhancing parameter
Property.
Secondly, passing through the reduction effect of second group of image:
The enhancing parameter obtained according to above-mentioned estimation restores image, calculates image and warp without enhancing processing
The peak value signal-to-noise ratio (Peak Signal to Noise Ratio, PSNR) of enhancing processing image is crossed, and is compared without enhancing
The image of processing and the PSNR for going back original image, data are as listed in table 3.As can be seen that image and reduction without enhancing processing
The PSNR of image is obviously improved compared with the image without enhancing processing with the PSNR (before reduction) by enhancing processing image,
Show that this programme can effectively restore original image.
In addition, Fig. 6 has chosen any two images in experiment, respectively and its original image is given from left to right, increased
Strong image and the corresponding intuitive contrast effect for going back original image are as shown in the figure.
Table 3
It should be understood that the size of the serial number of each step is not meant that the order of the execution order in above-described embodiment, each process
Execution sequence should be determined by its function and internal logic, the implementation process without coping with the embodiment of the present invention constitutes any limit
It is fixed.
In addition, the present embodiment first movement terminal, the second mobile terminal refer to a certain terminal, " first, second " herein only
It is not meant in specific implementation of the invention certain for the convenience stated and referred to being used to distinguish different terminals
Have corresponding first movement terminal, the second mobile terminal.
Corresponding to the detection method of image enhancement processing described in foregoing embodiments, Fig. 7 shows the embodiment of the present invention and mentions
The structural block diagram of the detection device of the image enhancement processing of confession.For ease of description, portion related to the present embodiment is illustrated only
Point.
Referring to Fig. 7, which includes:
First computing unit 71 calculates the quantized-dct DCT coefficient distribution period of described image；
Second computing unit 72 calculates the enhancing parameter Estimation of described image according to the quantization DCT coefficient distribution period
Value；
Judging unit 73, for judging whether the enhancing estimates of parameters is equal to 1；
Testing result output unit 74, if the enhancing estimates of parameters be equal to 1, described image be detected as without
The image of compression domain enhancing processing；If the enhancing estimates of parameters is not equal to 1, described image is detected as by compression domain
Enhance the image of processing.
Optionally, if the enhancing estimates of parameters is not equal to 1, described device further include:
Reduction unit, reciprocal using the enhancing estimates of parameters carry out compression domain enhancing processing to described image, also
Original goes out described image by the original image before compression domain enhancing processing.
Optionally, first computing unit 71 includes:
The picture element matrix of described image is divided into 8 × 8 blocks of pixels not overlapped by piecemeal subelement；
DCT handles subelement, carries out DCT processing to each blocks of pixels, obtain each blocks of pixels 8 ×
8 DCT coefficient matrix；
Subelement is extracted, frequency band B is extractednQuantization DCT coefficient sequenceN=i+j, 0≤n≤14；
First computation subunit calculates separately the quantization DCT coefficient sequenceGoal-selling function E (xl) letter
Numerical value, by finding the goal-selling function E (xl) obtain Local Minimum functional value position, determine the quantization DCT system
Number distribution period.
Optionally, first computing unit 71 further include:
Subelement is deformed, by the goal-selling function E (xl) deformation are as follows:
F(xl)=E (xl)-max{E(x1), E (x2) ... E (xl-1)}。
Optionally, second computing unit 72 includes:
Second computation subunit calculates the function of goal-selling function G (x) according to the quantization DCT coefficient distribution period
Value；
Subelement is obtained, the value of x when by the functional value minimum of the goal-selling function G (x) is retrieved as described image
Enhancing estimates of parameters.
It is apparent to those skilled in the art that for convenience of description and succinctly, only with above-mentioned each function
Can unit, module division progress for example, in practical application, can according to need and by above-mentioned function distribution by different
Functional unit, module are completed, i.e., the internal structure of described device is divided into different functional unit or module, more than completing
The all or part of function of description.Each functional unit in embodiment, module can integrate in one processing unit, can also
To be that each unit physically exists alone, can also be integrated in one unit with two or more units, it is above-mentioned integrated
Unit both can take the form of hardware realization, can also realize in the form of software functional units.In addition, each function list
Member, the specific name of module are also only for convenience of distinguishing each other, the protection scope being not intended to limit this application.Above system
The specific work process of middle unit, module, can refer to corresponding processes in the foregoing method embodiment, and details are not described herein.
Those of ordinary skill in the art may be aware that list described in conjunction with the examples disclosed in the embodiments of the present disclosure
Member and algorithm steps can be realized with the combination of electronic hardware or computer software and electronic hardware.These functions are actually
It is implemented in hardware or software, the specific application and design constraint depending on technical solution.Professional technician
Each specific application can be used different methods to achieve the described function, but this realization is it is not considered that exceed
The scope of the present invention.
In embodiment provided by the present invention, it should be understood that disclosed device and method can pass through others
Mode is realized.For example, system embodiment described above is only schematical, for example, the division of the module or unit,
Only a kind of logical function partition, there may be another division manner in actual implementation, such as multiple units or components can be with
In conjunction with or be desirably integrated into another system, or some features can be ignored or not executed.Another point, it is shown or discussed
Mutual coupling or direct-coupling or communication connection can be through some interfaces, the INDIRECT COUPLING of device or unit or
Communication connection can be electrical property, mechanical or other forms.
The unit as illustrated by the separation member may or may not be physically separated, aobvious as unit
The component shown may or may not be physical unit, it can and it is in one place, or may be distributed over multiple
In network unit.It can select some or all of unit therein according to the actual needs to realize the mesh of this embodiment scheme
's.
It, can also be in addition, the functional units in various embodiments of the present invention may be integrated into one processing unit
It is that each unit physically exists alone, can also be integrated in one unit with two or more units.Above-mentioned integrated list
Member both can take the form of hardware realization, can also realize in the form of software functional units.
If the integrated unit is realized in the form of SFU software functional unit and sells or use as independent product
When, it can store in a computer readable storage medium.Based on this understanding, the technical solution of the embodiment of the present invention
Substantially all or part of the part that contributes to existing technology or the technical solution can be with software product in other words
Form embody, which is stored in a storage medium, including some instructions use so that one
Computer equipment (can be personal computer, server or the network equipment etc.) or processor (processor) execute this hair
The all or part of the steps of bright each embodiment the method for embodiment.And storage medium above-mentioned include: USB flash disk, mobile hard disk,
Read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic
The various media that can store program code such as dish or CD.
Embodiment described above is merely illustrative of the technical solution of the present invention, rather than its limitations；Although referring to aforementioned reality
Applying example, invention is explained in detail, those skilled in the art should understand that: it still can be to aforementioned each
Technical solution documented by embodiment is modified or equivalent replacement of some of the technical features；And these are modified
Or replacement, the spirit and model of each embodiment technical solution of the embodiment of the present invention that it does not separate the essence of the corresponding technical solution
It encloses.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all in essence of the invention
Made any modifications, equivalent replacements, and improvements etc., should all be included in the protection scope of the present invention within mind and principle.
Claims (8)
1. a kind of detection method of image enhancement processing characterized by comprising
Calculate the quantized-dct DCT coefficient distribution period of given image；
According to the quantized-dct DCT coefficient distribution period, the enhancing estimates of parameters of described image is calculated；It is described
Enhancing estimates of parameters refers to the estimated value estimated the enhancing parameter of the compression domain of the given image, calculation method
Are as follows: the functional value of goal-selling function G (x) is calculated according to the quantized-dct DCT coefficient distribution period；It will be described
The value of x when the functional value minimum of goal-selling function G (x) is retrieved as the enhancing estimates of parameters of described image；
Judge whether the enhancing estimates of parameters is equal to 1；
If the enhancing estimates of parameters is equal to 1, described image is detected as enhancing the image of processing without compression domain；If
The enhancing estimates of parameters is not equal to 1, then described image is detected as enhancing the image of processing by compression domain.
2. the method as described in claim 1, which is characterized in that if the enhancing estimates of parameters is not equal to 1, the method is also
Include:
Compression domain enhancing processing is carried out to described image using the reciprocal of enhancing estimates of parameters, described image is restored and exists
By the original image before compression domain enhancing processing.
3. method according to claim 1 or 2, which is characterized in that the quantization DCT coefficient for calculating described image is distributed week
Phase includes:
The picture element matrix of described image is divided into 8 × 8 blocks of pixels not overlapped；
DCT processing is carried out to each blocks of pixels, obtains 8 × 8 DCT coefficient matrix of each blocks of pixels, it will
Described 8 × 8 DCT coefficient matrix is divided into 15 frequency bands, and n-th of frequency band is expressed as Bn；
Extract frequency band BnQuantization DCT coefficient sequenceWherein,Indicate that frequency is in enhanced DCT coefficient matrix
The quantization DCT coefficient of (i, j), n=i+j, 0≤n≤14；
Calculate separately the quantization DCT coefficient sequenceGoal-selling function E (xl) functional value, it is described pre- by finding
If objective function E (xl) obtain Local Minimum functional value position, determine the quantization DCT coefficient distribution period.
4. method as claimed in claim 3, which is characterized in that calculate separately the quantization DCT coefficient sequence described
Goal-selling function E (xl) functional value before, the method also includes:
By the goal-selling function E (xl) deformation are as follows:
F(xl)=E (xl)-max{E(x1),E(x2),…E(xl-1),
Wherein, the l is integer.
5. a kind of detection device of image enhancement processing characterized by comprising
First computing unit, for calculating the quantized-dct DCT coefficient distribution period of given image；
Second computing unit, for calculating the increasing of described image according to the quantized-dct DCT coefficient distribution period
Strong estimates of parameters；The enhancing estimates of parameters, which refers to, estimates the enhancing parameter of the compression domain of the given image
Estimated value；
Judging unit, for judging whether the enhancing estimates of parameters is equal to 1；
Testing result output unit, if being equal to 1 for the enhancing estimates of parameters, described image is detected as without over-voltage
The image of contracting domain enhancing processing；If the enhancing estimates of parameters is not equal to 1, described image is detected as increasing by compression domain
The image of strength reason；
Second computing unit includes:
Second computation subunit, for calculating the function of goal-selling function G (x) according to the quantization DCT coefficient distribution period
Value；
Subelement is obtained, the value of x when being used for the functional value minimum by the goal-selling function G (x) is retrieved as described image
Enhancing estimates of parameters.
6. device as claimed in claim 5, which is characterized in that if the enhancing estimates of parameters is not equal to 1, described device is also
Include:
Reduction unit, for carrying out compression domain enhancing processing to described image using the reciprocal of the enhancing estimates of parameters, also
Original goes out described image by the original image before compression domain enhancing processing.
7. such as device described in claim 5 or 6, which is characterized in that first computing unit includes:
Piecemeal subelement, for the picture element matrix of described image to be divided into 8 × 8 blocks of pixels not overlapped；
DCT handles subelement, for carrying out DCT processing to each blocks of pixels, obtain each blocks of pixels 8 ×
8 DCT coefficient matrix；Wherein, described 8 × 8 DCT coefficient matrix is divided into 15 frequency bands, and n-th of frequency band is expressed as Bn；
Subelement is extracted, for extracting frequency band BnQuantization DCT coefficient sequenceWherein,Indicate enhanced DCT system
Frequency is the quantization DCT coefficient of (i, j), n=i+j, 0≤n≤14 in matrix number；
First computation subunit, for calculating separately the quantization DCT coefficient sequenceGoal-selling function E (xl) letter
Numerical value, by finding the goal-selling function E (xl) obtain Local Minimum functional value position, determine the quantization DCT system
Number distribution period.
8. device as claimed in claim 7, which is characterized in that first computing unit further include:
Subelement is deformed, is used for the goal-selling function E (xl) deformation are as follows:
F(xl)=E (xl)-max{E(x1),E(x2),…E(xl-1),
Wherein, the l is integer.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510368803.7A CN106327470B (en) | 2015-06-29 | 2015-06-29 | A kind of detection method and device of image enhancement processing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510368803.7A CN106327470B (en) | 2015-06-29 | 2015-06-29 | A kind of detection method and device of image enhancement processing |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106327470A CN106327470A (en) | 2017-01-11 |
| CN106327470B true CN106327470B (en) | 2019-08-02 |
Family
ID=57721591
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510368803.7A Active CN106327470B (en) | 2015-06-29 | 2015-06-29 | A kind of detection method and device of image enhancement processing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106327470B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114598872B (en) * | 2020-12-07 | 2023-06-23 | 中国科学院深圳先进技术研究院 | JPEG compressed domain enhanced parameter estimation method, device, medium and computing equipment |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101572819A (en) * | 2009-06-03 | 2009-11-04 | 北京航空航天大学 | Reversible image watermark method based on quantized DCT coefficient zero values index |
| CN102236883A (en) * | 2010-04-27 | 2011-11-09 | 株式会社理光 | Image enhancing method and device as well as object detecting method and device |
| CN102724552A (en) * | 2012-05-30 | 2012-10-10 | 华为技术有限公司 | Image coding method, image decoding method and device |
| CN103310426A (en) * | 2012-03-14 | 2013-09-18 | 安凯(广州)微电子技术有限公司 | Image enhancement method and device based on joint photographic experts group (JPEG) frequency-domain transformation |
-
2015
- 2015-06-29 CN CN201510368803.7A patent/CN106327470B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101572819A (en) * | 2009-06-03 | 2009-11-04 | 北京航空航天大学 | Reversible image watermark method based on quantized DCT coefficient zero values index |
| CN102236883A (en) * | 2010-04-27 | 2011-11-09 | 株式会社理光 | Image enhancing method and device as well as object detecting method and device |
| CN103310426A (en) * | 2012-03-14 | 2013-09-18 | 安凯(广州)微电子技术有限公司 | Image enhancement method and device based on joint photographic experts group (JPEG) frequency-domain transformation |
| CN102724552A (en) * | 2012-05-30 | 2012-10-10 | 华为技术有限公司 | Image coding method, image decoding method and device |
Non-Patent Citations (4)
| Title |
|---|
| Image Enhancement Using a Contrast Measure in the Compressed Domain;Jinshan Tang 等;《IEEE SIGNAL PROCESSING LETTERS》;20031031;第10卷(第10期);第289-292页 * |
| 基于Retinex理论的压缩域图像增强方法研究;汪荣贵 等;《计算机研究与发展》;20111231;第48卷(第2期);第259-270页 * |
| 基于离散余弦变换的自适应图像增强;靳建峰 等;《武汉大学学报（工学版）》;20090831;第42卷(第4期);第532-536页 * |
| 离散余弦变换域内的Retinex图像增强模型研究;张新龙 等;《工程图学学报》;20111231(第6期);第33-42页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106327470A (en) | 2017-01-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zha et al. | From rank estimation to rank approximation: Rank residual constraint for image restoration | |
| Toderici et al. | Full resolution image compression with recurrent neural networks | |
| Wang et al. | D3: Deep dual-domain based fast restoration of JPEG-compressed images | |
| Foi et al. | Pointwise shape-adaptive DCT for high-quality denoising and deblocking of grayscale and color images | |
| Li et al. | An efficient deep convolutional neural networks model for compressed image deblocking | |
| KR20140136019A (en) | Method and apparatus for performing hierarchical super-resolution of an input image | |
| WO2013056129A1 (en) | Region-based image compression | |
| JP2006262278A (en) | Decoder, decoding method, and program therefor | |
| CN110383332B (en) | Image processing device, image processing method, and image processing program | |
| CN104392207A (en) | Characteristic encoding method for recognizing digital image content | |
| Wen et al. | Learning flipping and rotation invariant sparsifying transforms | |
| Bae et al. | DCT-QM: A DCT-based quality degradation metric for image quality optimization problems | |
| Douda et al. | A reduced domain pool based on DCT for a fast fractal image encoding | |
| CN106327470B (en) | A kind of detection method and device of image enhancement processing | |
| AU2002366676B2 (en) | Method, apparatus and software for lossy data compression and function estimation | |
| JP2004523941A (en) | Improving resolution by nearest-neighbor class filtering | |
| CN113256744B (en) | Image coding and decoding method and system | |
| Mukundan | Transform coding using discrete Tchebichef polynomials | |
| Liu et al. | Sparsity-based soft decoding of compressed images in transform domain | |
| CN111918061A (en) | Coding block dividing method and device, storage medium and electronic equipment | |
| Li | Inverse halftoning with nonlocal regularization | |
| Mukherjee et al. | Image filtering in the compressed domain | |
| Okarma et al. | A statistical reduced-reference approach to digital image quality assessment | |
| Chen et al. | A generic postprocessing technique for image compression | |
| Dar et al. | Image restoration via successive compression |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |