CN106530261A - Double-dynamic blurred image restoration method - Google Patents
Double-dynamic blurred image restoration method Download PDFInfo
- Publication number
- CN106530261A CN106530261A CN201611240091.1A CN201611240091A CN106530261A CN 106530261 A CN106530261 A CN 106530261A CN 201611240091 A CN201611240091 A CN 201611240091A CN 106530261 A CN106530261 A CN 106530261A
- Authority
- CN
- China
- Prior art keywords
- image
- layer
- current
- blurred
- kernel
- 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 31
- 230000009977 dual effect Effects 0.000 claims description 11
- 238000013507 mapping Methods 0.000 claims description 10
- 238000011478 gradient descent method Methods 0.000 claims description 6
- 238000005457 optimization Methods 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 6
- OAICVXFJPJFONN-UHFFFAOYSA-N Phosphorus Chemical compound [P] OAICVXFJPJFONN-UHFFFAOYSA-N 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- FFBHFFJDDLITSX-UHFFFAOYSA-N benzyl N-[2-hydroxy-4-(3-oxomorpholin-4-yl)phenyl]carbamate Chemical compound OC1=C(NC(=O)OCC2=CC=CC=C2)C=CC(=C1)N1CCOCC1=O FFBHFFJDDLITSX-UHFFFAOYSA-N 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000012804 iterative process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012806 monitoring device Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/73—Deblurring; Sharpening
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
The invention relates to a double-dynamic blurred image restoration method. The double-dynamic blurred image restoration method comprises the following steps of: obtaining a double-dynamic blurred image y, establishing an image pyramid of the double-dynamic blurred image, and establishing a motion blurred kernel and an out-of-focus blurred kernel of the double-dynamic blurred image y according to the image pyramid, wherein the resolution ratio of the blurred image in the image pyramid is sequentially increased from the top layer to the bottom layer; calculating optimal values of the motion blurred kernel, the out-of-focus blurred kernel and a clear image in a current layer of the image pyramid; and, according to the calculated optimal values of the motion blurred kernel, the out-of-focus blurred kernel and the clear image of the current image layer, calculating optimal values of the motion blurred kernel, the out-of-focus blurred kernel and the clear image of a next image layer till the clear image in the bottom layer of the image pyramid is calculated. By means of the method provided in the embodiment of the invention, the clear image is estimated in a process of continuously optimizing the motion blurred kernel and the out-of-focus blurred kernel; and thus, the clear image has enough definition.
Description
Technical Field
The invention relates to the technical field of image processing, in particular to a dual-dynamic blurred image restoration method.
Background
With the lapse of time, the antenna of the mobile electronic eye can go deep into each corner of the city, playing an increasingly important role and protecting the driving of our 'safe city'. The safe city is also an important embodiment for measuring the modernization management level of the city, and is an important measure for realizing the safety and stability of the city and even the whole country. Only by establishing a reasonable and effective urban video monitoring and management system, a government management part can find problems at the first time and provide countermeasures and emergency plans.
Monitoring devices in the urban video monitoring management system, for example, cameras, surveillance videos and the like often acquire blurred images, and in order to acquire clear images, in the prior art, the blurred images are processed through a blurring algorithm. For dual dynamic images, i.e. images acquired in a dual dynamic environment, for example, a camera on a moving vehicle acquires an image of another moving vehicle. In the process of obtaining a clear image by processing the double-dynamic image by using a deblurring algorithm, the clear image is usually estimated by constructing a single blur kernel for motion blur, for example, constructing a blur kernel for high-speed motion, vehicle shake, distance change and other factors.
However, in a dual dynamic environment, there are many reasons for image blur, and if a blur checking image is constructed for only one blur factor and processed, then deblurring processing cannot be implemented for other blur factors, so that the deblurred image cannot achieve a sufficient sharpness effect.
Disclosure of Invention
To overcome the problems in the related art, the present invention provides a dual dynamic blurred image restoration method.
According to a first aspect of embodiments of the present invention, there is provided a dual dynamic blurred image restoration method, including: acquiring a double-dynamic-blurred image y, establishing an image pyramid of the double-dynamic-blurred image, and establishing a motion blurred kernel and an out-of-focus blurred kernel of the double-dynamic-blurred image y according to the image pyramid, wherein the resolution of blurred images in the image pyramid is sequentially increased from a top layer to a bottom layer;
calculating the optimal values of a motion blur kernel, an out-of-focus blur kernel and a clear image of the current layer in the image pyramid;
and calculating the optimal values of the motion blur kernel, the out-of-focus blur kernel and the clear image of the next image layer according to the optimal values of the motion blur kernel, the out-of-focus blur kernel and the clear image calculated by the current image layer until the clear image of the bottom layer of the image pyramid is calculated.
Preferably, the calculating the optimal values of the motion blur kernel, the defocus blur kernel and the sharp image of the current layer in the image pyramid includes:
establishing a mapping relation between the double dynamic blurred images and the sharp images, wherein y is k1*k2X +, wherein k1Representing a motion blur kernel, k2Representing an out-of-focus blur kernel, x representing a sharp image, y representing a blurred image, and representing noise;
converting the mapping relation between the double dynamic fuzzy image and the clear image into a pair k1、k2And x, the optimization formula is as follows:wherein, β1、β2And λ is a weight value of the sum,representing the square of the norm of L2,wherein σ is weight, | · |. non-woven phosphor0The norm of L0 is shown,is a gradient operator;
will be provided withThe method is divided into two sub-formulas,and
k calculated from the previous layer1And k2By the formulaCalculating the optimal value of x of the current image layer;
according to the optimal value of x of the current image layer, calculating a formula by a gradient descent method:to obtain k of the current image layer1And k2The optimum value of (c).
Preferably, the formula is calculated according to the optimal value of x of the current image layer by a gradient descent method:to obtain k of the current image layer1And k2The optimal values of (a) include:
will be a formulaIs split intoAnd
blurring a kernel k for motion in gradient space1Performing estimation to obtain a formulaIs converted into
According to the optimal value of x of the current image layer and the current k2The optimum value of (1) is calculatedTo obtain k of the current image layer1The optimum value of (d);
according to the optimal value of x of the current image layer and the current k1The optimum value of (1) is calculatedTo obtain k of the current image layer2The optimum value of (c).
The technical scheme provided by the embodiment of the invention can have the following beneficial effects:
in the method provided by the embodiment of the invention, deblurring processing is carried out on two blurring factors, namely a motion blurring kernel and a defocusing blurring kernel of a blurred image. Establishing an image pyramid of the blurred image, sequentially estimating a clear image of each image layer from the top layer of the image pyramid, after estimating a motion blur kernel and an out-of-focus blur kernel of the current layer, iterating the motion blur kernel and the out-of-focus blur kernel in the next image layer, and estimating a clear image of the next image layer according to the motion blur kernel and the out-of-focus blur kernel. On the premise of continuously optimizing the motion blur kernel and the defocus blur kernel, the clear image of the next layer of image in the image pyramid is estimated, so that the clear image estimation optimization degree is increased, and the clear image is ensured to be finally obtained.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention, as claimed.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and together with the description, serve to explain the principles of the invention.
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without inventive exercise.
FIG. 1 is a schematic flow chart of a dual dynamic blurred image restoration method according to an embodiment of the present invention;
fig. 2 is a schematic flow chart of step S102 according to an embodiment of the present invention;
fig. 3 is a flowchart illustrating step S1025 according to an embodiment of the present invention.
Detailed Description
Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the invention, as detailed in the appended claims.
In a double-dynamic environment, the reasons for causing image blur are manifold, wherein two most important reasons are out-of-focus blur caused by the distance change of a motion blur kernel caused by motion.
Fig. 1 is a schematic flow chart of a dual dynamic blurred image restoration method according to an embodiment of the present invention.
In step S101, a dual dynamic blurred image y is obtained, an image pyramid of the dual dynamic blurred image is established, and a motion blurred kernel and an out-of-focus blurred kernel of the dual dynamic blurred image y are established according to the image pyramid, wherein the resolution of the blurred image in the image pyramid is sequentially increased from the top layer to the bottom layer.
Image pyramids are widely used in various visual applications. An image pyramid is a set of images y1,y2,......,yqAll images in the set are derived from the same original image, and the original image is continuously scaled until the desired pyramid level is obtained.
From the mathematical essence, the process of image blurring is generally understood as the process of convolving an original sharp image with a point spread function (also called a blurring kernel), and the influence of noise often exists. Then, the image deblurring technique is a deconvolution process. In the image deblurring process, if the blur kernel is known, the image restoration solves the estimation value of a real image according to the known blur kernel and the noise statistical characteristics. However, in general, since the blur kernel is unknown, it is necessary to estimate the blur kernel at the same time as estimating the sharp image.
In step S201, calculating optimal values of a motion blur kernel, a defocus blur kernel, and a sharp image of a current layer in an image pyramid;
and calculating optimal values of a motion blur kernel, a defocus blur kernel and a sharp image of each image layer from the top layer of the image pyramid. The specific calculation step, see fig. 2, is a schematic flow chart of step S102 provided in the embodiment of the present invention.
In step S1021, a mapping relationship between the double dynamic blurred image and the sharp image is established, where y is k1*k2X +, wherein k1Representing a motion blur kernel, k2Representing the out-of-focus blur kernel, x representing a sharp image, y representing a blurred image, and noise.
Knowing the sharp image x and the blurred image y, the mapping relationship between the blurred image and the sharp image can be expressed as: k is1*k2X + wherein k1Representing a motion blur kernel, k2Representing the out-of-focus blur kernel, x representing the sharp image, y representing the blurred image, and noise, (-) representing the convolution operation.
In step S1022, the mapping relationship between the double dynamic blurred image and the sharp image is converted into pair k1、k2And x, the optimization formula is as follows:wherein, β1、β2And λ is a weight value of the sum,representing the square of the norm of L2,wherein σ is weight, | · |. non-woven phosphor0The norm of L0 is shown,is a gradient operator.
In the above step, in the mapping relation between the clear image and the blurred image, y is known, and k is1、k2And x is unknown, so a sharp image cannot be calculated by the mapping relation, and therefore, the mapping relation needs to be converted into k in an iterative process1、k2And x.
The optimization formula is as follows:wherein, β1、β2And λ is a weight value of the sum,representing the square of the norm of L2,wherein σ is weight, | · |. non-woven phosphor0The norm of L0 is shown,is a gradient operator.
In step S1023, theThe method is divided into two sub-formulas,and
further, to the formulaSolving, splitting the solution into 2 formulas:
and
wherein, the formulaIn known k1、k2Then, the optimized value of x, formula, can be solvedWhen x is known, k can be solved1、k2The optimum value of (c).
In step S1024, k is calculated from the previous layer1And k2By the formulaThe optimal value of x for the current image layer is calculated.
When the current image layer estimates the optimal value of x, k calculated by the previous layer1And k2Optimization formula for iterating values into current image layer after amplifying according to image pyramid scaling ratioAnd calculating the optimal estimation value of the clear image x of the current image layer.
For an optimized estimation method of the sharp image x, see the following steps:
firstly, initializing parameters, and enabling x to be y;
1. order toLet mu be 2 lambda
2. Order to
3. Computing
4. Repeat 2, 4 until mu > mu, let mu be 2 mumax
5. The β is made to be 2 β, and 1 to 5 are repeated until β is more than βmax
Wherein F (-) is a fast Fourier transform, F-1(. cndot.) is an inverse fast Fourier transform,is a conjugate operation on the result of the Fourier transform, and
and sequentially iterating in each layer of image of the image pyramid until the optimal value of the clear image is obtained in the image layer at the bottom layer of the image pyramid.
In step S1025, according to the optimal value of x of the current image layer, the formula is calculated by the gradient descent method:to obtain k of the current image layer1And k2The optimum value of (c).
Substituting the optimal value of x calculated in the previous step into a formulaIn (1), calculating a motion blur kernel k1And defocus blur kernel k2And computing a motion blur kernel k1And defocus blur kernel k2And (5) substituting the stack into the next image layer to solve the clear image x.
Referring to fig. 3, a flow chart of step S1025 provided by the embodiment of the invention is shown, and in step S10251, a formula is shownIs split intoAnd
due to the formulaWhen a sharp image x is known, a motion blur kernel k is present1And defocus blur kernel k2Two unknowns, therefore, willSplitting intoAndwherein,for solving motion blur kernel k1，For solving out-of-focus fuzzy kernel k2。
In step S10252, the motion blur kernel k is processed in gradient space1Performing estimation to obtain a formulaIs converted into
Due to k1Is a motion blur kernel and thus can be estimated in gradient space, and thus, the formulaCan be rewritten as
The gradient descent method is used to solve the two equations in step S10252, here bySolving for k for example1。
Before treatment, forAndrespectively processed by soft threshold function to increaseAnddegree of sparsity of; the soft threshold function is as follows;
order toThe function J (k)1) Is a derivative ofTo find the optimum k1It is necessary to select a starting valueOrder toUp toOrTerminating the iteration to obtainIs k1The optimum value of (c). Where ρ is the step size, either constant or byThe calculation, η sum, is a positive constant equal to about 0 for the termination condition.
In step S10253, the optimal value of x according to the current image layer and the current k2The optimum value of (1) is calculatedTo obtain k of the current image layer1The optimum value of (c).
K calculated from the previous layer1And k2By the formulaAfter the optimal value of x of the current image layer is calculated, the optimal value of x and the current k are calculated2Substituting the optimal value into the formulaIn this way, k for the current image layer can be calculated1The optimum value of (c).
In step S10254, the optimal value of x according to the current image layer and the current k1The optimum value of (1) is calculatedTo obtain k of the current image layer2The optimum value of (c).
To k is paired2When estimating, x and y are first processed by a difference of gaussians function, i.e.Can be rewritten asWhere DoG (-) is a gaussian difference function. In pair k2Before estimation, the method also needs to process dog (x) and dog (y) through a soft threshold function, so as to increase the sparsity of dog (x) and dog (y).
In pair k2To carry outDuring estimation, specific frequency information, mainly edges and corners, is reserved through a Gaussian difference function, and then fuzzy kernel estimation is carried out.
K calculated from the previous layer1And k2By the formulaAfter the optimal value of x of the current image layer is calculated, the optimal value of x and the current k are calculated1Substituting the optimal value into the formulaObtaining the out-of-focus fuzzy kernel k of the current image layer2The optimum value of (c).
As can be seen from the above description, in the embodiment of the present invention, an image is processed by two blur factors, namely, a motion blur kernel and a defocus blur kernel, in a deblurring process, an image pyramid of a blurred image is established, a clear image is estimated from the top layer of the image pyramid, the motion blur kernel and the defocus blur kernel calculated in a current image layer are used as known conditions for estimating a clear image in a next layer, and then the clear image is continuously optimized until the last layer of the image pyramid obtains the clearest image.
Other embodiments of the invention will be apparent to those skilled in the art from consideration of the specification and practice of the disclosure herein. This application is intended to cover any variations, uses, or adaptations of the invention following, in general, the principles of the invention and including such departures from the present disclosure as come within known or customary practice within the art to which the invention pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the invention being indicated by the following claims.
It will be understood that the invention is not limited to the precise arrangements described above and shown in the drawings and that various modifications and changes may be made without departing from the scope thereof. The scope of the invention is limited only by the appended claims.
Claims (5)
1. A method for dual dynamic blurred image restoration, comprising:
acquiring a double-dynamic-blurred image y, establishing an image pyramid of the double-dynamic-blurred image, and establishing a motion blurred kernel and an out-of-focus blurred kernel of the double-dynamic-blurred image y according to the image pyramid, wherein the resolution of blurred images in the image pyramid is sequentially increased from a top layer to a bottom layer;
calculating the optimal values of a motion blur kernel, an out-of-focus blur kernel and a clear image of the current layer in the image pyramid;
and calculating the optimal values of the motion blur kernel, the out-of-focus blur kernel and the clear image of the next image layer according to the optimal values of the motion blur kernel, the out-of-focus blur kernel and the clear image calculated by the current image layer until the clear image of the bottom layer of the image pyramid is calculated.
2. The method of claim 1, wherein calculating optimal values of a motion blur kernel, a defocus blur kernel, and a sharp image for a current layer in an image pyramid comprises:
establishing a mapping relation between the double dynamic blurred images and the sharp images, wherein y is k1*k2X +, wherein k1Representing a motion blur kernel, k2Representing an out-of-focus blur kernel, x representing a sharp image, y representing a blurred image, and representing noise;
converting the mapping relation between the double dynamic fuzzy image and the clear image into a pair k1、k2And x, the optimization formula is as follows:wherein, β1、β2And λ is a weight value of the sum,representing the square of the norm of L2,wherein σ is weight, | · |. non-woven phosphor0The norm of L0 is shown,is a gradient operator;
will be provided withThe method is divided into two sub-formulas,and
k calculated from the previous layer1And k2By the formulaCalculating the optimal value of x of the current image layer;
according to the optimal value of x of the current image layer, calculating a formula by a gradient descent method:to obtain k of the current image layer1And k2The optimum value of (c).
3. The method according to claim 2, wherein the formula is calculated by a gradient descent method according to the optimal value of x of the current image layer:to obtain k of the current image layer1And k2The optimal values of (a) include:
will be a formulaIs split intoAnd
blurring a kernel k for motion in gradient space1Performing estimation to obtain a formulaIs converted into
According to the optimal value of x of the current image layer and the current k2The optimum value of (1) is calculatedTo obtain k of the current image layer1The optimum value of (d);
according to the optimal value of x of the current image layer and the current k1The optimum value of (1) is calculatedTo obtain k of the current image layer2The optimum value of (c).
4. Method according to claim 3, characterized in that the optimal value of x is determined from the current picture layer and the current k2The optimum value of (1) is calculatedBefore, comprising:
to pairAndrespectively processed by soft threshold function to increaseAnddegree of sparseness of.
5. A method as claimed in claim 3, characterized in that the maximum of x according to the current picture layerFigure of merit and current k1The optimum value of (1) is calculatedTo obtain k of the current image layer2The optimal values of (a) include:
processing x and y by using a Gaussian difference function, and converting the formulaIs converted intoWherein DoG (·) is a gaussian difference function;
processing the DoG (x) and the DoG (y) by using a soft threshold function to increase the sparsity of the DoG (x) and the DoG (y);
according to the optimal value of x of the current image layer and the current k1The optimum value of (1) is calculatedTo obtain k of the current image layer2The optimum value of (c).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611240091.1A CN106530261B (en) | 2016-12-28 | 2016-12-28 | A kind of double dynamic fuzzy image recovery methods |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611240091.1A CN106530261B (en) | 2016-12-28 | 2016-12-28 | A kind of double dynamic fuzzy image recovery methods |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106530261A true CN106530261A (en) | 2017-03-22 |
| CN106530261B CN106530261B (en) | 2019-03-19 |
Family
ID=58339139
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611240091.1A Active CN106530261B (en) | 2016-12-28 | 2016-12-28 | A kind of double dynamic fuzzy image recovery methods |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106530261B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107945127A (en) * | 2017-11-27 | 2018-04-20 | 南昌大学 | A kind of High-motion picture deblurring method based on image column gray probability uniformity |
| WO2018227465A1 (en) * | 2017-06-15 | 2018-12-20 | 中山大学 | Sparse positive source separation model-based image deblurring algorithm |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101968881A (en) * | 2010-10-27 | 2011-02-09 | 东南大学 | Motion blurring and defocusing composite blurring image restoration method |
| CN104331871A (en) * | 2014-12-02 | 2015-02-04 | 苏州大学 | Image de-blurring method and image de-blurring device |
| CN104599242A (en) * | 2014-12-09 | 2015-05-06 | 西安电子科技大学 | Multi-scale non-local regularization blurring kernel estimation method |
| CN105389820A (en) * | 2015-11-18 | 2016-03-09 | 成都中昊英孚科技有限公司 | Infrared image definition evaluating method based on cepstrum |
| CN105809642A (en) * | 2016-03-11 | 2016-07-27 | 中山大学 | Natural image blind motion blur elimination method based on L0 regularization |
| CN105957024A (en) * | 2016-04-20 | 2016-09-21 | 西安电子科技大学 | Blind deblurring method based on image block prior and sparse norm |
| CN106096605A (en) * | 2016-06-02 | 2016-11-09 | 史方 | A kind of image obscuring area detection method based on degree of depth study and device |
-
2016
- 2016-12-28 CN CN201611240091.1A patent/CN106530261B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101968881A (en) * | 2010-10-27 | 2011-02-09 | 东南大学 | Motion blurring and defocusing composite blurring image restoration method |
| CN104331871A (en) * | 2014-12-02 | 2015-02-04 | 苏州大学 | Image de-blurring method and image de-blurring device |
| CN104599242A (en) * | 2014-12-09 | 2015-05-06 | 西安电子科技大学 | Multi-scale non-local regularization blurring kernel estimation method |
| CN105389820A (en) * | 2015-11-18 | 2016-03-09 | 成都中昊英孚科技有限公司 | Infrared image definition evaluating method based on cepstrum |
| CN105809642A (en) * | 2016-03-11 | 2016-07-27 | 中山大学 | Natural image blind motion blur elimination method based on L0 regularization |
| CN105957024A (en) * | 2016-04-20 | 2016-09-21 | 西安电子科技大学 | Blind deblurring method based on image block prior and sparse norm |
| CN106096605A (en) * | 2016-06-02 | 2016-11-09 | 史方 | A kind of image obscuring area detection method based on degree of depth study and device |
Non-Patent Citations (1)
| Title |
|---|
| 谭海鹏等: "基于正则化约束的遥感图像多尺度去模糊", 《中国图象图形学报》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018227465A1 (en) * | 2017-06-15 | 2018-12-20 | 中山大学 | Sparse positive source separation model-based image deblurring algorithm |
| CN107945127A (en) * | 2017-11-27 | 2018-04-20 | 南昌大学 | A kind of High-motion picture deblurring method based on image column gray probability uniformity |
| CN107945127B (en) * | 2017-11-27 | 2021-05-07 | 南昌大学 | High-speed motion image deblurring method based on image column gray probability consistency |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106530261B (en) | 2019-03-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP7078139B2 (en) | Video stabilization methods and equipment, as well as non-temporary computer-readable media | |
| CN105469363B (en) | Method and apparatus based on blurred picture generation clear image | |
| EP3132418B1 (en) | Non local image denoising | |
| US8908989B2 (en) | Recursive conditional means image denoising | |
| Wang et al. | Improved scheme of estimating motion blur parameters for image restoration | |
| US20140294318A1 (en) | Gray image processing method and apparatus | |
| JP2017199235A (en) | Focus correction processing method by learning type algorithm | |
| KR20150032822A (en) | Method and apparatus for filtering an image | |
| CN108648162A (en) | A kind of gradient correlation TV factor graph picture denoising deblurring methods based on noise level | |
| US9202265B2 (en) | Point spread function cost function with non-uniform weights | |
| US20130343669A1 (en) | Blur-kernel estimation from spectral irregularities | |
| CN106530261B (en) | A kind of double dynamic fuzzy image recovery methods | |
| KR101921608B1 (en) | Apparatus and method for generating depth information | |
| CN110458767A (en) | A kind of blur ed image restoration method under poisson noise | |
| CN104504667B (en) | image processing method and device | |
| KR20170115757A (en) | Apparatus and method for generaing depth information | |
| TWI783721B (en) | Method for determining deblur filter coefficients | |
| CN109685839B (en) | Image alignment method, mobile terminal and computer storage medium | |
| Turgay et al. | Texture and edge preserving multiframe super‐resolution | |
| Conti et al. | A regularized deep learning approach for image de-blurring | |
| Patil et al. | Implementation of restoration of blurred image using blind deconvolution algorithm | |
| Morshed et al. | Frame by frame digital video denoising using multiplicative noise model | |
| Tsai et al. | An improved adaptive deconvolution algorithm for single image deblurring | |
| Murthy et al. | Performance evaluation of image restoration methods for comparative analysis with and without noise | |
| Dhanakshirur et al. | Evidence based feature selection and collaborative representation towards learning based PSF estimation for motion deblurring |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | 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 | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20211201 Address after: 241000 12th floor, advertising creative complex building, Wuhu advertising industrial park, middle Beijing Road, Jiujiang District, Wuhu City, Anhui Province Patentee after: CHANGXIN INTELLIGENT CONTROL NETWORK TECHNOLOGY CO.,LTD. Address before: 518057 1002, phase II, international student entrepreneurship building, No. 29, South Ring Road, gaoxinyuan, Nanshan District, Shenzhen, Guangdong Province Patentee before: TONGGUAN TECHNOLOGY (SHENZHEN) CO.,LTD. |
|
| TR01 | Transfer of patent right |