CN106709879A - Spatial variation point diffusion function smoothing method based on simple lens calculating imaging - Google Patents
Spatial variation point diffusion function smoothing method based on simple lens calculating imaging Download PDFInfo
- Publication number
- CN106709879A CN106709879A CN201611121879.0A CN201611121879A CN106709879A CN 106709879 A CN106709879 A CN 106709879A CN 201611121879 A CN201611121879 A CN 201611121879A CN 106709879 A CN106709879 A CN 106709879A
- Authority
- CN
- China
- Prior art keywords
- fuzzy core
- picture
- image
- point spread
- function
- 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 40
- 238000009499 grossing Methods 0.000 title claims abstract description 13
- 238000003384 imaging method Methods 0.000 title abstract description 7
- 238000009792 diffusion process Methods 0.000 title abstract description 3
- 230000008569 process Effects 0.000 claims abstract description 12
- 238000009826 distribution Methods 0.000 claims description 8
- 238000001914 filtration Methods 0.000 claims description 5
- 238000005315 distribution function Methods 0.000 claims description 3
- 230000000750 progressive effect Effects 0.000 claims description 3
- PCHJSUWPFVWCPO-UHFFFAOYSA-N gold Chemical compound [Au] PCHJSUWPFVWCPO-UHFFFAOYSA-N 0.000 claims 1
- 239000010931 gold Substances 0.000 claims 1
- 229910052737 gold Inorganic materials 0.000 claims 1
- 230000000694 effects Effects 0.000 abstract description 6
- 238000002474 experimental method Methods 0.000 abstract description 2
- 238000012545 processing Methods 0.000 abstract description 2
- 230000004075 alteration Effects 0.000 description 5
- 230000008859 change Effects 0.000 description 5
- 238000013461 design Methods 0.000 description 4
- 230000003287 optical effect Effects 0.000 description 4
- 230000009467 reduction Effects 0.000 description 3
- 238000011161 development Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 206010010071 Coma Diseases 0.000 description 1
- 238000002059 diagnostic imaging Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000009931 harmful effect Effects 0.000 description 1
- 229920013655 poly(bisphenol-A sulfone) Polymers 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012360 testing method Methods 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
- G06T5/70—Denoising; Smoothing
-
- 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
-
- 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
- G06T2207/10016—Video; Image sequence
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
The invention provides a spatial variation point diffusion function smoothing method based on simple lens calculating imaging. The method is characterized by fully using characteristics of sparsity, similarity and the like of a single lens fuzzy kernel; firstly, using a blind convolution algorithm to estimate a fuzzy kernel of each image block, and combining fuzzy kernel TV priori; and then carrying out smoothing and de-noising processing on the fuzzy kernel, and based on a filtered PSF of neighborhood information, replacing an existing PSF. Compared to the prior art, by using the method of the invention, a complex experiment environment and equipment are not needed, the spatial variational PSF can be estimated through combining a correlation algorithm based on a simple lens system; robustness of a deconvolution process is increased; and a noise, a ringing effect and the like are restrained.
Description
Technical field
Present invention relates generally to digital image processing field, refer in particular to a kind of space that picture is calculated as based on unzoned lens and become
Change PSF (point spread function) smoothing method.
Background technology
In modern optical system, picture quality can be reduced because of optical parallax, and most with spherical mirror knot
Single convex lens of structure can all be influenceed by such as aberration, spherical aberration, coma etc..In order to solve this predicament, existing optical imagery
System is mainly by complicated combined lens to make up the aberration of single eyeglass, and the camera lens of such as slr camera may include tens of
Individual independent single eyeglass or lens set.But, the design of complex combination camera lens while image quality is improved, undoubtedly also significantly
The cost of lens design manufacture is increased, and the volume and weight of camera lens is consequently increased.Therefore, how simple mirror is being eliminated
Piece group aberration, it is ensured that while image quality, reduces the design and manufacture cost of camera lens, makes it lighter, be future optical into
As the development trend of system.Unzoned lens system has a potential prospect in many scientific domains, such as unmanned plane, remotely sensed image with
And medical imaging.
In recent years, with the fast development of image restoration technology, the method such as image deblurring is more and more ripe, in camera lens certain
A little aberrations and the eyeglass of Modified geometrical distortion of eliminating can calculate camera work replacement by deblurring etc., therefore, unzoned lens imaging
The combination of (as shown in Figure 2) and calculating camera work is increasingly becoming a new research direction.
The key of image restoration is that accurate PSF (point spread function) estimates that accurate PSF can be in image deconvolution mistake
The generation of journey suppressed ringing phenomenon.Generally it is considered its distribution with its place PSF (point spread function) in method before
Locus it is unrelated, whole blurred picture is a fuzzy core for the overall situation, i.e. space invariance PSF；But in actual reality
Test in result, we can see that PSF is changed with change in location, i.e. spatial variations PSF.
In PSF estimations, optimal problem iterative is generally used.The fuzzy kernel estimates of unzoned lens are typically in many chis
Carried out in degree space, set up one by roughly to fine image pyramid.The ratio of each two image layer is and with 3 × 3
Gaussian function or delta function as fuzzy core initial value, the number of plies of image layer determines by the size of fuzzy core.Pass through
The iteration in different levels metric space progressively tries to achieve final preferable fuzzy core successively, and in each layer of metric space,
The fuzzy core that will be tried to achieve in last layer subdimension space first tries to achieve potential clear figure as initial value with reference to blurred picture
Picture, then using potential picture rich in detail and fuzzy core as known terms, then obtains picture rich in detail.
It is if it is assumed that being that space invariance PSF can cause to obscure the problems of kernel estimates in current unzoned lens imaging
PSF estimates inaccurate, the quality of the harmful effects such as ringing effect reduction result images can be produced in deconvolution process, if false
If the PSF of spatial variations, otherwise existing method needs sufficiently expensive, fine experimental situation and equipment, otherwise need complexity
Iterative algorithm design, spend substantial amounts of operation time.And because image is by piecemeal, compared to using whole blurred picture come
Estimate PSF, be used for estimating that the pixel of PSF is also accordingly reduced after piecemeal, this may cause to estimate the accuracy and stability of PSF
Drop is fixed.Therefore, the characteristic for being calculated as picture according to unzoned lens proposes that a robustness and the spatial variations PSF being easily achieved are smoothed
Method is that simple lens is calculated as urgent problem.
The content of the invention
It is inaccurate for the fuzzy kernel estimates of spatial variations in existing unzoned lens imaging method, calculating process take it is oversize,
The problems such as complex experiment environment is difficult to, it is an object of the invention to provide a kind of space change that picture is calculated as based on unzoned lens
Change point spread function smoothing method, the features such as this method makes full use of the openness and similitude of simple lens fuzzy core, profit first
The fuzzy core of each image block is estimated with blind convolution algorithm, with reference to fuzzy core TV priori, then by fuzzy core carry out it is smooth and
Denoising, existing PSF is replaced based on the filtered PSF of neighborhood information.Compared to existing method, the method need not complexity
Experimental situation and equipment, can just estimate the PSF of spatial variations based on unzoned lens system combination related algorithm, and increase is gone
The robustness of convolution process, suppresses noise and ringing effect etc..
The technical scheme is that,
A kind of spatial variations point spread function smoothing method that picture is calculated as based on unzoned lens, is comprised the following steps,
S1. blurred picture is obtained using unzoned lens combination web camera fuselage, if the size of the blurred picture for obtaining
It is H × W pixels.
S2. the point spread function estimation problem that unzoned lens is calculated as in is converted into blind convolved image and restores problem,
The fuzzy core of unzoned lens is obtained by blind convolved image restoration algorithm, is comprised the following steps that：
S21. blurred picture is divided into M × N block images according to size, each image block size isPixel；Profit
The corresponding fuzzy core of each image block is estimated with blind convolved image restoration algorithm.
S2.2 using the similitude of the adjacent fuzzy core of unzoned lens and openness, be smoothed using adjacent area and
Noise in threshold function table removal point spread function is set, fuzzy core P (u, v) of unzoned lens is obtained；
By each is estimated in S21 current image block and its point spread function array of adjacent image block by with low pass
Wave filter convolution algorithm reduces random error, and filtered result replaces the original point spread function of current image block, wherein
The structure of low pass filter is as follows：
Denoising is carried out after filtering, by setting the threshold function table related to locus successively by each image block
Value in point spread function less than threshold value is set to zero：
Wherein P (u, v) refers to that p (u, v) is distributed by point spread function final after denoising, and p (u, v) is at filtering
The distribution function of the current point spread function after reason before denoising, (u, v) represents the certain point in P (u, v), threshold value T (u, v)
For：
T (u, v)=1-H (u) H (v)
H (u) functions and H (v) same distributions, it is defined as follows：
Wherein, R is the radius of fuzzy core, and parameter alpha is the intensity for controlling denoising function to remove noise.
Further, the corresponding fuzzy core of each image block is estimated using blind convolution in step S21 of the present invention, wherein blind
The object function of the fuzzy nuclear issue of the solution of convolution can be expressed as：
Wherein, k represents the fuzzy core of unzoned lens, also known as point spread function；X represents picture rich in detail；Y is obtained in representing S1
Blurred picture；Represent convolution operation；Section 1It is data fit term, it is clear after expression blurred picture and convolution
The matching degree of clear image；Section 2It is the business of the norm with two norms of x, is the bound term to picture rich in detail x；3rd
Item μ | | k | |1The total variant prioris to fuzzy core k are represented, is the bound term to fuzzy core k；λ and μ are data
The weight coefficient of fit term and bound term；It is the conservation of energy to fuzzy core and nonnegativity restriction.
In the blind the solution of convolution fuzzy cores of step S21, by iterative formula (1), wherein iterative formula (1)
Process is carried out in multiscale space, and detailed process is as follows：
A. a picture size is set up by roughly to fine image pyramid, the ratio of each two image layer based on down-sampled
Example be
B. using 3 × 3 Gaussian function or delta function as the initial value of fuzzy core, the corresponding figure of current image block
As the number of plies of pyramidal image layer is determined by the size of the fuzzy core for setting with the ratio of the size of initial fuzzy core.
C. by the way that using IRLS, (interative least square method is in the different levels metric space of image pyramid successively
The ripe algorithm of this area) solution formula (1), progressive alternate tries to achieve final fuzzy core, in each layer of metric space, first general
The fuzzy core tried to achieve in last layer subdimension space is tried to achieve the y that blurred picture is substituted into formula (1) potential as initial value
Picture rich in detail x, then substitute into next metric space using potential picture rich in detail and fuzzy core as known terms, then obtain fuzzy
Core and picture rich in detail, until last metric space.
Advantageous Effects of the invention：
The present invention calculates imaging system based on unzoned lens, the estimation of unzoned lens PSF is converted into blind convolved image and is answered
Former algorithm, reduces in spatial variations PSF estimation procedures because image block causes the PSF stability estimated, proposes to be based on
Unzoned lens space-variant blur core is openness and PSF smoothing methods of neighborhood similarity, is estimated according to blind convolution algorithm
The PSFs of current and adjacent area image block, involves threshold function table and removes the robust that noise and increase PSF estimate by low pass filtered
Property, this method can well recover the details and suppressed ringing effect of image during blur image restoration.
Brief description of the drawings
Fig. 1 is the fuzzy core of the corresponding spatial variations of unzoned lens；
Fig. 2 is unzoned lens imaging system principle schematic；
Fig. 3 is based on unzoned lens spatial variations PSF smoothing method flow charts；
Fig. 4 is unzoned lens structural representation
Fig. 5 is that image block fuzzy core filters schematic diagram；
Fig. 6 is the final PSF for trying to achieve.
Specific embodiment
To make the object, technical solutions and advantages of the present invention clearer, below in conjunction with accompanying drawing to embodiment party of the present invention
Formula is described in further detail.
As shown in figure 3, a kind of spatial variations point spread function that picture is calculated as based on unzoned lens that the present embodiment is provided
Smoothing method, comprises the following steps：
Step one：Unzoned lens combination web camera fuselage obtains blurred picture, as shown in figure 4, institute in the present embodiment
The unzoned lens of use includes 3 simple lenses.The web camera that the present embodiment is used can shoot 1080 × 1920 pixels
Image, its photo-sensitive cell is 1/1.9 " CMOS of inch.
Step 2：The point spread function estimation problem that unzoned lens is calculated as in is converted into blind convolved image recovery to ask
Topic, i.e., obtain the fuzzy core of unzoned lens by blind convolved image restoration algorithm, and method is as follows：
Blurred picture is divided into 6 × 10 image blocks by S2.1, and each image block is 180 × 192 pixels, estimates each
The corresponding fuzzy core of image block.Estimate that the algorithm used by the fuzzy core of each image block proposes blind convolved image for Krishnan
Restoration algorithm combination fuzzy core priori, the object function that blind convolved image restores solution fuzzy core can be expressed as：
Wherein, k represents the fuzzy core of unzoned lens, also known as point spread function (PSF)；X represents picture rich in detail；Y represents mould
Paste image；Represent convolution operation；Section 1It is data fit term, clearly schemes after representing blurred picture and convolution
The matching degree of picture；Section 2It is the business of the norm with two norms of x, is the bound term to picture rich in detail x；Section 3 μ |
|k||1Total variant (TV) priori to fuzzy core k is represented, is the bound term to fuzzy core k, the mould for often using
Paste core priori is Gaussian prior, but the distribution of point diffusion is not simple Gauss or disk distribution, so TV priori is more
Plus the method for robust, while can be good at eliminating the noise of fuzzy core and ensure convergence；λ and μ be data fit term with about
The weight coefficient of beam；It is the conservation of energy to fuzzy core and nonnegativity restriction.
In blind the solution of convolution fuzzy core, by iterative formula (1), the process of wherein iterative formula (1) is
Carried out in multiscale space, detailed process is as follows：
A. a picture size is set up by roughly to fine image pyramid, the ratio of each two image layer based on down-sampled
Example be
B. using 3 × 3 Gaussian function or delta function as the initial value of fuzzy core, the corresponding figure of current image block
As the number of plies of pyramidal image layer is determined by the size of the fuzzy core for setting with the ratio of the size of initial fuzzy core；
C. by utilizing interative least square method solution formula in the different levels metric space of image pyramid successively
(1), progressive alternate tries to achieve final fuzzy core, in each layer of metric space, will be tried to achieve in last layer subdimension space first
Fuzzy core as initial value, the y that blurred picture is substituted into formula (1) is tried to achieve into potential picture rich in detail x, then potential clear
Clear image and fuzzy core substitute into next metric space as known terms, then obtain fuzzy core and picture rich in detail, until last
Metric space.
Because image is by piecemeal, compared to PSF is estimated using whole blurred picture, it is used for estimating the picture of PSF after piecemeal
Vegetarian refreshments is also accordingly reduced, and this may cause to estimate accuracy and the stability reduction of PSF.Make full use of unzoned lens adjacent fuzzy
The similitude of core and openness, is smoothed and set using adjacent area threshold function table and can be very good to remove PSF's
Noise and increase PSF estimation robustness so that effectively suppress due to PSF estimate it is inaccurate caused by deconvolution process
The ringing effect of middle appearance.
Fuzzy core that S21 is estimated as shown in figure 1, each point represents the fuzzy core corresponding to each image block, by Fig. 1
It can be seen that PSF major part region value be zero and without centre symmetry, distribution be not simple Gaussian Profile or
Disk is distributed, and the shape of fuzzy core is changed by disk to slender type.
S2.2 by each is estimated in S21 current image block and its point spread function array of adjacent image block by with
Low pass filter convolution algorithm reduces random error, and filtered result replaces the original point spread function of current image block,
As shown in figure 5, the structure of wherein low pass filter is as follows：
In 3 × 3 wave filter, the weight in the region nearer apart from current image block is bigger.Because image is by piecemeal, phase
Compared with PSF is estimated using whole blurred picture, it is used for estimating that the pixel of PSF is also accordingly reduced after piecemeal, this may cause
Estimate accuracy and the stability reduction of PSF, some noises may be produced in the borderline region of the PSF for estimating.It is ensuing
Step is to remove these noises, by setting the threshold function table related to locus successively by the point spread function of each image block
Value in number less than threshold value is set to zero, and threshold value is larger in PSF central areas, and more remote apart from PSF centers, threshold value is gradually reduced.This
Sample is to meet the regularity of distribution of PSF while also been removed the noise on PSF borders, finally give unzoned lens fuzzy core P (u,
v)：
Wherein P (u, v) refers to that p (u, v) is distributed by point spread function final after denoising, and p (u, v) is at filtering
The distribution function of the current point spread function after reason before denoising, (u, v) represents the certain point in P (u, v), threshold value T (u, v)
For：
T (u, v)=1-H (u) H (v)
H (u) functions and H (v) same distributions, it is defined as follows：
Wherein, R is the radius of fuzzy core, and parameter alpha is the intensity for controlling denoising function to remove noise.The space for finally giving
Change PSF as shown in fig. 6, eliminate most of noise before compared to smooth.
The explanation of the preferred embodiment of the present invention contained above, this be in order to describe technical characteristic of the invention in detail, and
Be not intended to be limited in the content of the invention in the concrete form described by embodiment, carry out according to present invention purport other
Modification and modification are also protected by this patent.The purport of present invention is to be defined by the claims, rather than by embodiment
Specific descriptions are defined.
Claims (3)
1. a kind of spatial variations point spread function smoothing method that picture is calculated as based on unzoned lens, it is characterised in that：Including with
Lower step,
S1. blurred picture is obtained using unzoned lens combination web camera fuselage, if the size of the blurred picture for obtaining be H ×
W pixels；
S2. the point spread function estimation problem that unzoned lens is calculated as in is converted into blind convolved image and restores problem, that is, led to
The fuzzy core that blind convolved image restoration algorithm obtains unzoned lens is crossed, is comprised the following steps that：
S21. blurred picture is divided into M × N block images according to size, each image block size isPixel；Using blind
Convolved image restoration algorithm estimates the corresponding fuzzy core of each image block；
S2.2 is smoothed and is set using adjacent area using the similitude of the adjacent fuzzy core of unzoned lens and openness
Noise in threshold function table removal point spread function, obtains fuzzy core P (u, v) of unzoned lens；
By each is estimated in S21 current image block and its point spread function array of adjacent image block by with LPF
Device convolution algorithm reduces random error, and filtered result replaces the original point spread function of current image block, wherein low pass
The structure of wave filter is as follows：
Denoising is carried out after filtering, is successively expanded the point of each image block by setting the threshold function table related to locus
In scattered function zero is set to less than the value of threshold value：
Wherein P (u, v) refers to p (u, v) to be distributed by point spread function final after denoising, after p (u, v) is filtering process
The distribution function of the current point spread function before denoising, (u, v) represents the certain point in P (u, v), and threshold value T (u, v) is：
T (u, v)=1-H (u) H (v)
H (u) functions and H (v) same distributions, it is defined as follows：
Wherein, R is the radius of fuzzy core, and parameter alpha is the intensity for controlling denoising function to remove noise.
2. the spatial variations point spread function smoothing method that picture is calculated as based on unzoned lens according to claim 1, its
It is characterised by：The corresponding fuzzy core of each image block is estimated using blind convolution in step S21, wherein blind the solution of convolution fuzzy core
The object function of problem can be expressed as：
Wherein, k represents the fuzzy core of unzoned lens, also known as point spread function；X represents picture rich in detail；Y represents the mould obtained in S1
Paste image；Represent convolution operation；Section 1It is data fit term, clearly schemes after representing blurred picture and convolution
The matching degree of picture；Section 2It is the business of the norm with two norms of x, is the bound term to picture rich in detail x；Section 3 μ |
|k||1The total variant prioris to fuzzy core k are represented, is the bound term to fuzzy core k；λ and μ are data fittings
With the weight coefficient of bound term；It is the conservation of energy to fuzzy core and nonnegativity restriction.
3. the spatial variations point spread function smoothing method that picture is calculated as based on unzoned lens according to claim 2, its
It is characterised by：In the blind the solution of convolution fuzzy cores of step S21, by iterative formula (1), wherein iterative formula (1)
Process carried out in multiscale space, detailed process is as follows：
A. a picture size is set up by roughly to fine image pyramid based on down-sampled, the ratio of each two image layer is
B. using 3 × 3 Gaussian function or delta function as the initial value of fuzzy core, the corresponding image gold of current image block
The number of plies of the image layer of word tower is determined by the size of the fuzzy core for setting with the ratio of the size of initial fuzzy core；
C. by utilizing interative least square method solution formula (1) in the different levels metric space of image pyramid successively,
Progressive alternate tries to achieve final fuzzy core, in each layer of metric space, the mould that will be tried to achieve in last layer subdimension space first
The y that blurred picture is substituted into formula (1) is tried to achieve potential picture rich in detail x, then potential clear figure by paste core as initial value
Picture and fuzzy core substitute into next metric space as known terms, then obtain fuzzy core and picture rich in detail, until last yardstick
Space.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611121879.0A CN106709879B (en) | 2016-12-08 | 2016-12-08 | A kind of spatial variations point spread function smoothing method that picture is calculated as based on unzoned lens |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611121879.0A CN106709879B (en) | 2016-12-08 | 2016-12-08 | A kind of spatial variations point spread function smoothing method that picture is calculated as based on unzoned lens |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106709879A true CN106709879A (en) | 2017-05-24 |
| CN106709879B CN106709879B (en) | 2017-12-15 |
Family
ID=58936380
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611121879.0A Active CN106709879B (en) | 2016-12-08 | 2016-12-08 | A kind of spatial variations point spread function smoothing method that picture is calculated as based on unzoned lens |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106709879B (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107369135A (en) * | 2017-06-22 | 2017-11-21 | 广西大学 | A kind of micro imaging system three-dimensional point spread function space size choosing method based on Scale invariant features transform algorithm |
| CN107590790A (en) * | 2017-09-21 | 2018-01-16 | 长沙全度影像科技有限公司 | A kind of unzoned lens fringe region deblurring method based on symmetrical edge filling |
| CN107622477A (en) * | 2017-08-08 | 2018-01-23 | 成都精工华耀机械制造有限公司 | A kind of RGBW images joint demosaicing and deblurring method |
| CN107680062A (en) * | 2017-10-12 | 2018-02-09 | 长沙全度影像科技有限公司 | A kind of micro- burnt Restoration method of blurred image based on l1/l2 priori combination Gaussian priors |
| CN107749051A (en) * | 2017-10-17 | 2018-03-02 | 长沙全度影像科技有限公司 | A kind of unzoned lens space-variant blur core smoothing method based on mean filter |
| CN108074221A (en) * | 2017-12-19 | 2018-05-25 | 长沙全度影像科技有限公司 | A kind of parametrization unzoned lens PSF methods of estimation |
| CN108765509A (en) * | 2018-05-22 | 2018-11-06 | 西南交通大学 | A kind of fast image reconstruction method for linear imaging system |
| CN109102476A (en) * | 2018-08-28 | 2018-12-28 | 北京理工大学 | A kind of multispectral image defocusing blurring kernel estimates method based on blur circle fitting |
| CN110475068A (en) * | 2019-08-30 | 2019-11-19 | 北京迈格威科技有限公司 | Image processing method and device |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104809706A (en) * | 2015-05-05 | 2015-07-29 | 中国人民解放军国防科学技术大学 | Single lens computational imaging method based on gentle image color change priori |
| US20160259758A1 (en) * | 2015-03-04 | 2016-09-08 | Animesh Khemka | Digital Camera Simulation Using Spatially Varying PSF Convolutions |
-
2016
- 2016-12-08 CN CN201611121879.0A patent/CN106709879B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160259758A1 (en) * | 2015-03-04 | 2016-09-08 | Animesh Khemka | Digital Camera Simulation Using Spatially Varying PSF Convolutions |
| CN104809706A (en) * | 2015-05-05 | 2015-07-29 | 中国人民解放军国防科学技术大学 | Single lens computational imaging method based on gentle image color change priori |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107369135A (en) * | 2017-06-22 | 2017-11-21 | 广西大学 | A kind of micro imaging system three-dimensional point spread function space size choosing method based on Scale invariant features transform algorithm |
| CN107622477A (en) * | 2017-08-08 | 2018-01-23 | 成都精工华耀机械制造有限公司 | A kind of RGBW images joint demosaicing and deblurring method |
| CN107590790A (en) * | 2017-09-21 | 2018-01-16 | 长沙全度影像科技有限公司 | A kind of unzoned lens fringe region deblurring method based on symmetrical edge filling |
| CN107680062A (en) * | 2017-10-12 | 2018-02-09 | 长沙全度影像科技有限公司 | A kind of micro- burnt Restoration method of blurred image based on l1/l2 priori combination Gaussian priors |
| CN107749051A (en) * | 2017-10-17 | 2018-03-02 | 长沙全度影像科技有限公司 | A kind of unzoned lens space-variant blur core smoothing method based on mean filter |
| CN108074221A (en) * | 2017-12-19 | 2018-05-25 | 长沙全度影像科技有限公司 | A kind of parametrization unzoned lens PSF methods of estimation |
| CN108765509A (en) * | 2018-05-22 | 2018-11-06 | 西南交通大学 | A kind of fast image reconstruction method for linear imaging system |
| CN108765509B (en) * | 2018-05-22 | 2020-08-14 | 西南交通大学 | Rapid image reconstruction method for linear imaging system |
| CN109102476A (en) * | 2018-08-28 | 2018-12-28 | 北京理工大学 | A kind of multispectral image defocusing blurring kernel estimates method based on blur circle fitting |
| CN109102476B (en) * | 2018-08-28 | 2021-06-08 | 北京理工大学 | Multispectral image defocusing fuzzy kernel estimation method based on circle of confusion fitting |
| CN110475068A (en) * | 2019-08-30 | 2019-11-19 | 北京迈格威科技有限公司 | Image processing method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106709879B (en) | 2017-12-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106709879B (en) | A kind of spatial variations point spread function smoothing method that picture is calculated as based on unzoned lens | |
| WO2020087607A1 (en) | Bi-skip-net-based image deblurring method | |
| CN111028177B (en) | Edge-based deep learning image motion blur removing method | |
| CN108122197B (en) | Image super-resolution reconstruction method based on deep learning | |
| CN101639938B (en) | Image restoration method based on double-edge wave filter and margin deconvolution | |
| CN102576454B (en) | Image deblurring using a spatial image prior | |
| CN111091503A (en) | Image out-of-focus blur removing method based on deep learning | |
| CN110796616B (en) | Turbulence degradation image recovery method based on norm constraint and self-adaptive weighted gradient | |
| Shen et al. | Convolutional neural pyramid for image processing | |
| CN111161360A (en) | Retinex theory-based image defogging method for end-to-end network | |
| CN112241939B (en) | Multi-scale and non-local-based light rain removal method | |
| CN106504207A (en) | A kind of image processing method | |
| CN114693558A (en) | Image Moire removing method and system based on progressive fusion multi-scale strategy | |
| CN111932480A (en) | Deblurred video recovery method and device, terminal equipment and storage medium | |
| Zhang et al. | Deep motion blur removal using noisy/blurry image pairs | |
| CN109300092A (en) | A kind of remote sensing images deblurring method based on dark | |
| CN106027854A (en) | United filtering denoising method which is applied to a camera and is applicable to be realized in FPGA (Field Programmable Gate Array) | |
| CN115937794A (en) | Small target object detection method and device, electronic equipment and storage medium | |
| Zou et al. | Joint wavelet sub-bands guided network for single image super-resolution | |
| Thakur et al. | Gradient and multi scale feature inspired deep blind gaussian denoiser | |
| CN112132757B (en) | General image restoration method based on neural network | |
| Guo et al. | Image blind deblurring using an adaptive patch prior | |
| CN110738624B (en) | Area-adaptive image defogging system and method | |
| CN108230251A (en) | Combined type image recovery method and device | |
| CN116523790A (en) | SAR image denoising optimization method, system and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |