CN110443246A - A kind of fish eye images effective coverage extracting method based on the optimization of equal value difference - Google Patents
A kind of fish eye images effective coverage extracting method based on the optimization of equal value difference Download PDFInfo
- Publication number
- CN110443246A CN110443246A CN201910747553.6A CN201910747553A CN110443246A CN 110443246 A CN110443246 A CN 110443246A CN 201910747553 A CN201910747553 A CN 201910747553A CN 110443246 A CN110443246 A CN 110443246A
- Authority
- CN
- China
- Prior art keywords
- eye images
- fish eye
- ellipse
- effective coverage
- oval
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
The present invention discloses a kind of fish eye images effective coverage extracting method based on the optimization of equal value difference, fish eye images effective coverage is indicated by establishing model of ellipse, and optimize ellipse by maximizing the mean value inside fish eye images effective coverage and the mean value of outside minimizing, and calculate the minimum value that external mean value subtracts internal mean value, to find global optimum, corresponding elliptic parameter is exported again, extracts fish eye images effective coverage image.The present invention has good robustness to picture material variation in noise spot outside the edge blurry of fish eye images effective coverage, effective coverage and effective coverage, improves the precision of fish eye images synthesis.
Description
Technical field
The present invention relates to technical field of image processing, in particular to a kind of fish eye images effective district based on the optimization of equal value difference
Domain extracting method.
Background technique
In the field of virtual reality of computer vision, scene walkthrough technology is fast-developing.And use fish eye lens
It obtains spherical panorama and is important one of scene walkthrough method.In spherical panorama map generalization, to obtain fish-eye image with most
The one-to-one relationship of whole spherical panorama image pixel, first has to the center of circle, the radius that provide fish-eye image image-region.Also to have
The field angle of the corresponding realistic space of this fish eye images.Currently there are many papers all and there is no accurate fish-eye image circle
Heart radius just carries out next step splicing, brings difficulty to next step splicing operation.
The center of circle radius and field angle for having directly given fish eye images in the prior art do not say it is how to come.In
In panorama picture formation based on the rotation of single flake, the coordinate of image and radius are considered as unknown number, led to during image mosaic
The optimizing index of splicing effect is crossed to adjust this parameter.Because there are many parameter for influencing image mosaic, if cannot give in advance
The higher image central coordinate of circle of precision and radius parameter out, then calculation will be greatlyd improve when carrying out optimal parameter search in the later period
Method computation complexity, and it is easy to cause algorithm is unstable can not converge to locally optimal solution, so that it is effective to influence fish eye images
The extraction in region.
Chinese patent (notification number: CN105678729B) discloses a kind of fish eye lens Panorama Mosaic method, passes through
A plurality of straight line is established to come in the optimization algorithm that the perspective view of fish eye images is input data to fish-eye imaging model parameter
It optimizes, the center of circle of fish-eye image circle and radius, nonlinear distortion are all its Optimal Parameters.But because fail that flake is determined in advance
The center of circle of figure and radius make fish eye lens splicing complexity greatly increase and be easy unstable.Therefore, panoramic mosaic is being carried out
Or before distortion parameter calibration, determine that the center of circle of fish eye images and radius are very necessary in advance.
There are mainly four types of the methods that current fish-eye image effective coverage is extracted, the first is the shape by seeking its binary image
The heart and area find out the center of circle and the radius of fish-eye image effective coverage.This method is simple and efficient, but is had inside image pure
It is easy to be not allowed when black region, and being more easier the case where occurring is the adjacent outside in effective coverage it is possible that a small amount of bright
Block, these bright blocks will have a direct impact on result precision.Second is scanning Beam Method.Using four or eight scan lines from image
Most edge toward medial movement, current scan line scanning just stops moving to the pixel that brightness is higher than certain threshold value, then again will
Scan line seeks round parameter as the tangent line of circle.Although this method can not by internal black regional effect, if
There are black color lump or outside to have bright block still can be impacted inside the edge of the circle.The third is the fish eye images based on region growing
Contours extract algorithm, this method first carry out region growing after by image binaryzation, then extract the progress of its marginal information
Least square fitting.Its region growing depends on the edge of luminance block, can fill up all interior in growth course
Portion black hole and the external part being recessed back are filled up.This method equally to effective coverage outside a small amount of luminance block incapability be
Power can even make luminance block become more serious, and fill up also imperfect to the part being recessed back, and it is inaccurate to eventually lead to result
Really.4th kind is the effective coverage for adding hough-circle transform, least-square fitting approach to obtain based on binaryzation.This method is
Image has been carried out binaryzation by one step, however the case where once encounter edge blurry, the result edge that binaryzation comes out will be curved
Bending is bent, keeps Hough loop truss algorithm entirely ineffective.Image, which cannot be handled, there are also it the case where elliptical distortion.Even if it will suddenly
Husband's circle transformation has been changed to Hough elliptic transformation, since Hough elliptic transformation computation complexity itself is high, it will make calculating speed
It is very slow, bring efficiency to perplex to industrialized batch detection.
Summary of the invention
Splice unstable problem for fish eye images in the prior art, the present invention proposes a kind of based on the optimization of equal value difference
Fish eye images effective coverage extracting method.
To achieve the goals above, the present invention the following technical schemes are provided:
A kind of fish eye images effective coverage extracting method based on the optimization of equal value difference, comprising the following steps:
S1: fisheye camera photographic subjects object is used, to obtain the original fish eye images of object；
S2: oval by buildingModel extract the effective coverages of original fish eye images；
S3: two-dimensional function is adjustedParameter make ellipseIt is overlapped, and leads to original fish eye images inner ellipse
It crosses model and calculates ellipseThe value e of corresponding objective functioni；
S4: will be ovalThe center of circle, major and minor axis add and subtract first time step-length respectively, can respectively obtain 16 kinds variation after
Ellipse setAnd oval set is calculated separately out by modelCorresponding objective function
Value set { ei+1, and work as min ei+1<min ei, repeat step S4；If min ei+1>min ei, then S5 is entered step；If
min ei+1With min eiIt is equal, then enter step S6；
First time step-length: being re-set as 0.8 times of size of first time step-length by S5, repeats step S4；
S6: setting 1/10th of original fish eye images width for step-length s ', will be ovalThe center of circle, major and minor axis
Ellipse after plus-minus step sizes s ' can be changed respectivelyAnd calculate separately out ellipseIt is corresponding
Objective function value set { e 'n, if min e 'n<min ei, then it represents that jump out local optimum success, i.e. min e 'nIt is
One better local optimum, then enter step S7；If min e 'n>min ei, then it represents that local optimum failure is jumped out,
That is min eiIt is exactly global optimum, enters step S8,
S7: being restored to first time step sizes for step-length s ', repeat step S4, S5, S6, after when ellipse variation m times
Obtain ellipseAnd min e 'm>min em, i.e. min emIt is global optimum, enters step S8,0≤i < n < m,
I, n, m are integer, indicate the number of oval variation；
S8: output is ovalParameter information, to export original fish eye images effective coverage image.
Preferably, the objective function e expression formula of the model of ellipse is following formula:
In formula (1),It indicatesNormalized function, I indicates original fish eye images, and x, y respectively indicate original fish
The abscissa and ordinate of any pixel point, c in eye image I1It is original fish eye images I in ellipseInternal mean value, c2
It is image I in ellipseExternal mean value；
It is a two-dimensional function, it is positive value on the inside of elliptic contour, is negative value on the outside of elliptic contour, and contain four
It is a with it is ovalRelated parameter Cx,Cy, a, b, then the expression formula of two-dimensional function is as follows:
In formula (2), CxIndicate center of circle abscissa, CyIndicate that center of circle ordinate, a indicate that elliptical long axis, b indicate oval
Short axle, in order to willIts range is normalized and adjusts, use is with minor function pairIt is handled:
The meaning of formula (3) is that a two-dimensional function is generated according to elliptic parameter, this two-dimensional function can describe image
Each pixel is inside ellipse or outside ellipse；Work asWhen, value 1 indicates pixel in elliptic wheel
It is wide internal；WhenWhen, value 0 indicates pixel outside elliptic contour；WhenWhen, value 0.5 indicates
Pixel is on elliptic contour boundary.
Preferably, the Luminance Distribution of the original fish eye images effective coverage is in the section of [a, b], and 2a > b.
Preferably, in the step S4, the first time step sizes are set as 1.
Preferably, in the step S4, the equal expression | min en-min en-1| < ε, ε are a preset values.
In conclusion by adopting the above-described technical solution, compared with prior art, the present invention at least has beneficial below
Effect:
The present invention is by finding the method that the oval internal mean value of fish eye images maximizes and oval external mean value minimizes
One ellipse expresses the content of fish eye images effective coverage；The present invention outside the edge blurry of effective coverage, effective coverage to making an uproar
Picture material variation has good robustness in sound point and effective coverage, corrects the side of providing for further fish eye images
Just.
It is extracted using effective coverage of the present invention to fish eye images, stability is high, and to camera lens elliptical distortion, edge
There are these situations of a small amount of hot spot not fail outside fuzzy, effective coverage, and has hot spot outside edge blurry, effective coverage
In the case of, it is smaller to extract error.
Detailed description of the invention:
Fig. 1 is to be extracted according to a kind of fish eye images effective coverage based on the optimization of equal value difference of exemplary embodiment of the present
Method flow schematic diagram.
Specific embodiment
Below with reference to embodiment and specific embodiment, the present invention is described in further detail.But this should not be understood
It is all that this is belonged to based on the technology that the content of present invention is realized for the scope of the above subject matter of the present invention is limited to the following embodiments
The range of invention.
In the description of the present invention, it is to be understood that, term " longitudinal direction ", " transverse direction ", "upper", "lower", "front", "rear",
The orientation or positional relationship of the instructions such as "left", "right", "vertical", "horizontal", "top", "bottom" "inner", "outside" is based on attached drawing institute
The orientation or positional relationship shown, is merely for convenience of description of the present invention and simplification of the description, rather than the dress of indication or suggestion meaning
It sets or element must have a particular orientation, be constructed and operated in a specific orientation, therefore should not be understood as to limit of the invention
System.
In the description of the present invention, unless otherwise specified and limited, it should be noted that term " installation ", " connected ",
" connection " shall be understood in a broad sense, for example, it may be mechanical connection or electrical connection, the connection being also possible to inside two elements can
, can also indirectly connected through an intermediary, for the ordinary skill in the art to be to be connected directly, it can basis
Concrete condition understands the concrete meaning of above-mentioned term.
Referring to Fig.1, the present invention provide it is a kind of based on equal value difference optimization fish eye images effective coverage extracting method, including with
Lower step:
S1: fisheye camera photographic subjects object is used, to obtain the original fish eye images of object.
In the present embodiment, certain brightness is needed to have when shooting to object using fisheye camera, to guarantee to clap
The original fish eye images taken the photograph can be identified.If the Luminance Distribution of original fish eye images effective coverage in the section of [a, b],
Then 2a > b.
S2: the effective coverage of original fish eye images is extracted by building model of ellipse.
In original fish eye images (for rectangular image), there are effective coverages and inactive area.It is reality in effective coverage
The light in the world maps the visual region that comes by camera lens；And inactive area is that light extraneous under normal circumstances all reflects not
To region, black is presented, is possible to have a small amount of hot spot in the case where subject is light source and light source is extremely strong.Cause
This need to extract the effective coverage of original fish eye images using a kind of method, avoid be by the influence of inactive area factor
Subsequent image procossing provides basis.
The effective coverage of ideally original fish eye images is a circle, and video camera is exactly to set according to being projected out a circle
Meter, but generated tangential abnormal since lens itself are not parallel with camera sensor plane (imaging plane) or the plane of delineation
Become, so that effective coverage becomes an ellipse.
Therefore the extraction problem of effective coverage is described as finding an ellipse by the present inventionTo represent original flake
The effective coverage of image, the present invention establish transverse and longitudinal coordinate system by origin of the bottom left vertex of original flake rectangular image, define mesh
Scalar functions e is oval to correspond to statementFind the minimum value of oval corresponding objective function, i.e. global optimum；Such as
Function is successional, therefore has many peak valleys, and it is optimal in its neighborhood that local optimum, which just looks like a peak valley, but this
A not all peak valley of peak valley is minimum, and global optimum is that minimum of all peak valleys.If jumping out local optimum,
Illustrate to have found peak valley more lower than this peak valley, can not jump and illustrate that this local optimum is also global optimum.
The expression formula of objective function e is as follows:
In formula (1),Indicate that a two-dimensional function, x, y respectively indicate any pixel point in original fish eye images I
Abscissa and ordinate, I indicates original fish eye images, c1It is image I in ellipseInternal mean value, c2It is image I
In ellipseExternal mean value.
It is a two-dimensional function, it is positive value on the inside of elliptic contour, is negative value on the outside of elliptic contour, that is, meetsAll coordinate x, y all on oval boundary, meetAll coordinate x, y all outside ellipse,
MeetAll coordinate x, y is inside ellipse；And containing there are four with ellipseRelated parameter Cx,Cy,
A, b, then the expression formula of two-dimensional function is as follows:
In formula (2), CxIndicate oval center of circle abscissa, CyIndicate that oval center of circle ordinate, a indicate ellipse along horizontal seat
Target zoom factor, that is, elliptical long axis, b indicate oval zoom factor, that is, elliptical short axle along ordinate.
In the present embodiment, in order to incite somebody to actionIts range is normalized and adjusts, use is with minor function pairIt is handled:
The meaning of formula (3) is that a two-dimensional function is generated according to elliptic parameter, this two-dimensional function can describe image
Each pixel is inside ellipse or outside ellipse；Work asWhen, value 1 indicates pixel in elliptic wheel
It is wide internal；WhenWhen, value 0 indicates pixel outside elliptic contour；WhenWhen, value 0.5 indicates
Pixel is on elliptic contour boundary；And there is smooth transition in elliptic contour edge.
S3: two-dimensional function is adjustedParameter make ellipseIt is overlapped, and counts with original fish eye images inner ellipse
It calculates ovalThe value e of corresponding objective functioni。
In the present embodiment, original fish eye images I is rectangular image, therefore can have an inner ellipse, when original flake
When the size of image I is fixed, then the center location of inner ellipse and length shaft size are all fixed, therefore the present invention needs
By adjusting two-dimensional functionParameter, such as the abscissa C in the center of circlex, ordinate Cy, long axis a, short axle b, make corresponding ellipseIt is overlapped with image inner ellipse.
To ellipseIt is oval after being overlapped with original fish eye images I inner ellipseCentral coordinate of circle and length
Axis parameter can fix, and calculate ellipse by formula (1), (2), (3)The value e of corresponding objective functioni。
S4: will be ovalThe center of circle, major and minor axis add and subtract first time step-length respectively, can respectively obtain 16 kinds variation after
Ellipse setAnd oval set is calculated separately out by modelCorresponding objective function
Value set { ei+1, and take fresh target functional minimum value min ei+1With eiIt compares.I indicates the number of oval variation,
That is i >=0, and be integer.
As min ei+1<min ei, repeat step S4；If min ei+1>min ei, then S5 is entered step；If min ei+1With
min eiIt is equal, then enter step S6.
In the present embodiment, settable first time step sizes s1It is 1, it will be ovalThe center of circle abscissa Cx, it is vertical
Coordinate Cy, long axis a, short axle b add and subtract first time step sizes s respectively1Afterwards, variation can be calculated by formula (1), (2), (3)
Ellipse afterwardsThe value of corresponding fresh target function, such as CxReduce, CyReducing, a reduces, and b reduction is a kind of combination,
CxReduce, CyReduce, a reduces, and b increase is another combination, four parameters a total of 24=16 kinds of combinations can be obtained 16 newly
The set e of the value of fresh target function can be obtained in the value of objective function1；Take set e1Middle minimum value mine1With e0It compares.
If min e1<e0, then make minimum value min e1Corresponding 4 Parameters variations come into force, thus after obtaining variation 1 time
It is oval4 parameters, such as minimum value min e1Corresponding 4 parameters are C 'x=Cx-s1, C 'y=Cy+s1, a '=
a-s1, b '=b-s1, then oval4 parameters be respectively new center of circle abscissa C 'x, ordinate C 'y, it is long axis a ', short
Axis b '.
It will be ovalCenter of circle abscissa C 'x, ordinate C 'y, long axis a ', that short axle b ' adds and subtracts initial step length respectively is big
Small s1Ellipse after being changedSet { the e of the value of fresh target function can be obtained2, if min e2<min e1, weight
Multiple step S4；If min e2>min e1, then S5 is entered step；If min e2=min e1, then S6 is entered step.
In the present embodiment, because the variation of function causes functional value essentially equal, therefore can be by min en=min
en-1It is interpreted as | min en-min en-1| < ε, ε are a default minimums of setting, to avoid there is unlimited repetitive cycling
Situation.
S5: keep ovalSecond of step-length is set 0.8 times of first time step-length by parameter constant, repeats to walk
Rapid S4.
In the present embodiment, step sizes are set as 0.8 times of size of last step sizes, i.e. s next timek=
0.8sk-1, k >=2, k are positive integer, such as s2=0.8s1=0.8.
S6: 1/10th of original fish eye images I width w are set by step-length s ', i.e.,And it will be ovalThe center of circle, major and minor axis add and subtract step sizes s ' respectively and can be changed after ellipseAnd it calculates separately out
It is ovalValue set { the e ' of corresponding objective functionn, if min e 'n<min ei, then it represents that jump out local optimum
It is worth successfully, i.e. min e 'nIt is a better local optimum, then enters step S7；If min e 'n>min ei, then it represents that it jumps
Local optimum fails out, i.e. min eiIt is exactly global optimum, enters step S8.
S7: step-length s ' is restored to first time step sizes s1, repeat step S4, S5, S6, the ellipse computed repeatedly every time
All be ellipse corresponding to target function value minimum after Parameters variation, until when it is oval change m (0≤i < n < m) it is secondary after obtain
It is ovalAnd min e 'm>min em, i.e. min emIt is exactly global optimum, enters step S8；
S8: output is ovalParameter information, extract in original fish eye images ovalThe image for including
Content, that is, the original fish eye images effective coverage image extracted.
In the present embodiment, when determining min emAfter being exactly global optimum, then corresponding ellipseCentral coordinate of circle and
Major and minor axis size is fixed, therefore can draw out ellipse on original fish eye images according to the center of circle and major and minor axis parameterIt removes inactive area, exports the image of the effective coverage of original fish eye images I, i.e., it is ovalInclude
Image.
Involved each module and unit are logic module and logic unit in present embodiment, in practical application
In, a logic unit can be a physical unit, be also possible to a part of a physical unit, can also be with multiple objects
The combination for managing unit is realized.In addition, in order to protrude innovative part of the invention, it will not be with this hair of solution in present embodiment
The technical issues of bright proposed, the less close unit of relationship introduced, but this does not indicate that there is no others in present embodiment
Unit.
It will be understood by those skilled in the art that the respective embodiments described above are to realize specific embodiments of the present invention,
And in practical applications, can to it, various changes can be made in the form and details, without departing from the spirit and scope of the present invention.
Claims (5)
1. a kind of fish eye images effective coverage extracting method based on the optimization of equal value difference, which is characterized in that including
S1: fisheye camera photographic subjects object is used, to obtain the original fish eye images of object；
S2: oval by buildingModel extract the effective coverages of original fish eye images；
S3: two-dimensional function is adjustedParameter make ellipseIt is overlapped with original fish eye images inner ellipse, and passes through mould
Type calculates ovalThe value e of corresponding objective functioni；
S4: will be ovalThe center of circle, major and minor axis add and subtract first time step-length respectively, can respectively obtain 16 kinds variation after ellipse
SetAnd oval set is calculated separately out by modelThe value of corresponding objective function
Gather { ei+1, and work as min ei+1<min ei, repeat step S4；If min ei+1>min ei, then S5 is entered step；If min
ei+1With min eiIt is equal, then enter step S6；
First time step-length: being re-set as 0.8 times of size of first time step-length by S5, repeats step S4；
S6: setting 1/10th of original fish eye images width for step-length s ', will be ovalThe center of circle, major and minor axis difference
Plus-minus step sizes s ' can be changed after ellipseAnd calculate separately out ellipseCorresponding mesh
Value set { the e ' of scalar functionsn, if min e 'n<min ei, then it represents that jump out local optimum success, i.e. min e 'nIt is one
Better local optimum, then enter step S7；If min e 'n>min ei, then it represents that jump out local optimum failure, i.e. min
eiIt is exactly global optimum, enters step S8,
S7: being restored to first time step sizes for step-length s ', repeat step S4, S5, S6, obtains until after ellipse variation m times
It is ovalAnd min e 'm>min em, i.e. min emIt is global optimum, enters step S8,0≤i < n < m, i, n, m
For integer, the number of oval variation is indicated；
S8: output is ovalParameter information, to export original fish eye images effective coverage image.
2. a kind of fish eye images effective coverage extracting method based on the optimization of equal value difference as described in claim 1, feature exist
In the objective function e expression formula of the model of ellipse is following formula:
E=c2-c1
In formula (1),It indicatesNormalized function, I indicates original fish eye images, and x, y respectively indicate original fish-eye image
As the abscissa and ordinate of any pixel point in I, c1It is original fish eye images I in ellipseInternal mean value, c2It is figure
As I is in ellipseExternal mean value；
A two-dimensional function, it on the inside of elliptic contour be positive value, on the outside of elliptic contour be negative value, and containing there are four with
It is ovalRelated parameter Cx,Cy, a, b, then the expression formula of two-dimensional function is as follows:
In formula (2), CxIndicate center of circle abscissa, CyIndicate that center of circle ordinate, a indicate that elliptical long axis, b indicate elliptical short
Axis, in order to incite somebody to actionIts range is normalized and adjusts, use is with minor function pairIt is handled:
The meaning of formula (3) is that a two-dimensional function is generated according to elliptic parameter, and it is each that this two-dimensional function can describe image
A pixel is inside ellipse or outside ellipse；Work asWhen, value 1 indicates pixel in elliptic contour
Portion；WhenWhen, value 0 indicates pixel outside elliptic contour；WhenWhen, value 0.5 indicates pixel
Point is on elliptic contour boundary.
3. a kind of fish eye images effective coverage extracting method based on the optimization of equal value difference as described in claim 1, feature exist
In the Luminance Distribution of the original fish eye images effective coverage is in the section of [a, b], and 2a > b.
4. a kind of fish eye images effective coverage extracting method based on the optimization of equal value difference as described in claim 1, feature exist
In in the step S4, the first time step sizes are set as 1.
5. a kind of fish eye images effective coverage extracting method based on the optimization of equal value difference as described in claim 1, feature exist
In, in the step S4, the equal expression | min en-min en-1| < ε, ε are a preset values.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910747553.6A CN110443246B (en) | 2019-08-14 | 2019-08-14 | Mean difference optimization-based fisheye image effective region extraction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910747553.6A CN110443246B (en) | 2019-08-14 | 2019-08-14 | Mean difference optimization-based fisheye image effective region extraction method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110443246A true CN110443246A (en) | 2019-11-12 |
| CN110443246B CN110443246B (en) | 2022-02-22 |
Family
ID=68435361
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910747553.6A Active CN110443246B (en) | 2019-08-14 | 2019-08-14 | Mean difference optimization-based fisheye image effective region extraction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110443246B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111179184A (en) * | 2019-11-29 | 2020-05-19 | 广东工业大学 | Fish-eye image effective region extraction method based on random sampling consistency |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101739547A (en) * | 2008-11-21 | 2010-06-16 | 中国科学院沈阳自动化研究所 | Precise identification and position method of robust coding point in image under complex background |
| CN103268592A (en) * | 2013-04-24 | 2013-08-28 | 南京邮电大学 | Method for correcting fisheye images |
| CN104167002A (en) * | 2014-08-28 | 2014-11-26 | 中国矿业大学 | Method for extracting effective area of fisheye image |
| KR20160040330A (en) * | 2014-07-30 | 2016-04-14 | 동의대학교 산학협력단 | A method of correcting for distorted image from fish-eye lens by using concentric circles type standard patterns |
| CN106651758A (en) * | 2016-12-16 | 2017-05-10 | 深圳市保千里电子有限公司 | Noisy fisheye image-based effective region extraction method and system |
| CN106910173A (en) * | 2017-02-27 | 2017-06-30 | 武汉大学 | The method that flake video wicket real time roaming is realized based on correcting fisheye image |
| CN106952218A (en) * | 2017-03-08 | 2017-07-14 | 南京杰迈视讯科技有限公司 | A kind of extracting method of panorama picture of fisheye lens effective coverage |
| CN107689033A (en) * | 2017-07-21 | 2018-02-13 | 哈尔滨工程大学 | A kind of fish eye images distortion correction method based on ellipse segmentation |
| CN109859276A (en) * | 2019-01-21 | 2019-06-07 | 桂林电子科技大学 | A kind of fish eye lens inner parameter self-calibrating method |
| CN110060259A (en) * | 2019-04-04 | 2019-07-26 | 东华大学 | A kind of fish eye lens effective coverage extracting method based on Hough transformation |
-
2019
- 2019-08-14 CN CN201910747553.6A patent/CN110443246B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101739547A (en) * | 2008-11-21 | 2010-06-16 | 中国科学院沈阳自动化研究所 | Precise identification and position method of robust coding point in image under complex background |
| CN103268592A (en) * | 2013-04-24 | 2013-08-28 | 南京邮电大学 | Method for correcting fisheye images |
| KR20160040330A (en) * | 2014-07-30 | 2016-04-14 | 동의대학교 산학협력단 | A method of correcting for distorted image from fish-eye lens by using concentric circles type standard patterns |
| CN104167002A (en) * | 2014-08-28 | 2014-11-26 | 中国矿业大学 | Method for extracting effective area of fisheye image |
| CN106651758A (en) * | 2016-12-16 | 2017-05-10 | 深圳市保千里电子有限公司 | Noisy fisheye image-based effective region extraction method and system |
| CN106910173A (en) * | 2017-02-27 | 2017-06-30 | 武汉大学 | The method that flake video wicket real time roaming is realized based on correcting fisheye image |
| CN106952218A (en) * | 2017-03-08 | 2017-07-14 | 南京杰迈视讯科技有限公司 | A kind of extracting method of panorama picture of fisheye lens effective coverage |
| CN107689033A (en) * | 2017-07-21 | 2018-02-13 | 哈尔滨工程大学 | A kind of fish eye images distortion correction method based on ellipse segmentation |
| CN109859276A (en) * | 2019-01-21 | 2019-06-07 | 桂林电子科技大学 | A kind of fish eye lens inner parameter self-calibrating method |
| CN110060259A (en) * | 2019-04-04 | 2019-07-26 | 东华大学 | A kind of fish eye lens effective coverage extracting method based on Hough transformation |
Non-Patent Citations (8)
| Title |
|---|
| CHUNYI CHEN等: ""Robust contour extraction of fisheye images for image-based virtual reality"", 《2010 3RD INTERNATIONAL CONGRESS ON IMAGE AND SIGNAL PROCESSING》 * |
| LUMINITA A.VESE等: ""A Multiphase Level Set Framework for Image Segmentation Using the Mumford and Shah Model"", 《INTERNATIONAL JOURNAL OF COMPUTER VISION》 * |
| SHENHAI ZHENG等: ""A variational approach to liver segmentation using statistics from multiple sources"", 《PHYSICS IN MEDICINE & BIOLOGY》 * |
| SHENHAI ZHENG等: ""B-Spline based globally optimal segmentation combining low-level and high-level information"", 《PATTERN RECOGNITION》 * |
| TONY F.CHAN等: ""Active contours without edges"", 《IEEE TRANSACTIONS ON IMAGE PROCESSING》 * |
| 常丹华等: ""提取鱼眼图像轮廓的算法改进及图像校正"", 《激光与红外》 * |
| 赵丹: ""鱼眼图像变形算法研究"", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
| 金敢峰: ""基于多特征融合的鱼眼图像行人检测与跟踪"", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111179184A (en) * | 2019-11-29 | 2020-05-19 | 广东工业大学 | Fish-eye image effective region extraction method based on random sampling consistency |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110443246B (en) | 2022-02-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP2678824B1 (en) | Determining model parameters based on transforming a model of an object | |
| CN105957015B (en) | A kind of 360 degree of panorama mosaic methods of threaded barrel inner wall image and system | |
| CN109903331B (en) | Convolutional neural network target detection method based on RGB-D camera | |
| CN102509093B (en) | Close-range digital certificate information acquisition system | |
| CN106683139A (en) | Fisheye-camera calibration system based on genetic algorithm and image distortion correction method thereof | |
| CN108537782B (en) | Building image matching and fusing method based on contour extraction | |
| CN104778656B (en) | Fisheye image correcting method based on spherical perspective projection | |
| US11403739B2 (en) | Methods and apparatus for retargeting and prioritized interpolation of lens profiles | |
| US20130307966A1 (en) | Depth measurement apparatus, image pickup apparatus, and depth measurement program | |
| JP2017005389A (en) | Image recognition device, image recognition method, and program | |
| WO2004055735A1 (en) | Pattern identification method, device thereof, and program thereof | |
| JP2010041417A (en) | Image processing unit, image processing method, image processing program, and imaging apparatus | |
| CN108230242B (en) | Method for converting panoramic laser point cloud into video stream | |
| BR112021005139A2 (en) | image processing apparatus and method, and, program for causing an image processing apparatus to perform an image process | |
| CN110021055A (en) | The method and its image processing apparatus and system of generation disparity map | |
| CN113391644A (en) | Unmanned aerial vehicle shooting distance semi-automatic optimization method based on image information entropy | |
| CN114268736A (en) | Tower footing dome camera shooting method with high space coverage | |
| US7006706B2 (en) | Imaging apparatuses, mosaic image compositing methods, video stitching methods and edgemap generation methods | |
| WO2022217834A1 (en) | Method and system for image processing | |
| CN110443246A (en) | A kind of fish eye images effective coverage extracting method based on the optimization of equal value difference | |
| CN110084754A (en) | A kind of image superimposing method based on improvement SIFT feature point matching algorithm | |
| WO2021170051A1 (en) | Digital photogrammetry method, electronic device, and system | |
| CN113436130A (en) | Intelligent sensing system and device for unstructured light field | |
| CN107590790B (en) | Simple lens edge area deblurring method based on symmetric edge filling | |
| CN116132729A (en) | Panoramic video stitching method and system for landslide monitoring |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |