CN104751458B - A kind of demarcation angular-point detection method based on 180 ° of rotation operators - Google Patents
A kind of demarcation angular-point detection method based on 180 ° of rotation operators Download PDFInfo
- Publication number
- CN104751458B CN104751458B CN201510127733.6A CN201510127733A CN104751458B CN 104751458 B CN104751458 B CN 104751458B CN 201510127733 A CN201510127733 A CN 201510127733A CN 104751458 B CN104751458 B CN 104751458B
- Authority
- CN
- China
- Prior art keywords
- mrow
- rotation
- point
- template
- image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses a kind of demarcation angular-point detection method based on 180 ° of rotation operators, comprise the steps：(1) pending uncalibrated image I is obtained, (2) judge grey scale change degree Z1；(3) 180 ° of rotation registration Z2 are judged；(4) angle point of image is extracted.The present invention is during camera calibration, according to the rotational symmetry of the X-comers surrounding pattern in three faces of calibrating block, designs a kind of template based on 180 ° of rotation operators, realizes the efficient detection to calibrating block angle point.Pass through experimental data and interpretation of result, the detection of this method angle steel joint is with very high accuracy and stability, general scaling board image is can be applied not only to, the calibrating block image in multiple faces is also applied for, for having certain gradient and the uncalibrated image of distortion equally to have preferable effect.
Description
Technical field
The present invention relates to the image processing field in surface-mounting equipment, more particularly to one kind is entered based on 180 ° of rotation operators
The method of angle point in the quick accurate detection camera calibration of row.
Background technology
Camera calibration is the important component in machine vision and image procossing, and its precision is determined under many circumstances
The performance of whole three-dimension measuring system.In precise electronic manufacture series equipment, surface-mounting equipment is for being carried out to PCB
The equipment of component attachment.Therefore, the importance of change system accuracy is self-evident.Generally, it is to utilize chess
Disk table images are as uncalibrated image, and angle point therein is exactly the point that brightness of image change differs greatly with adjoint point intensity of variation.Angle
Point detection is the important step of camera calibration.By extracting the corner location information on image, it may be determined that three-dimensional measurement system
In system, the relation of the information of the every bit of three-dimensional body between image plane coordinate and space coordinates, so as to obtain video camera
Inside and outside parameter, and then determine its locus.
The approach of camera calibration is, according to camera model, the mould of video camera to be solved by the image coordinate of known features point
Shape parameter.Every pixel is all obtained by transmission projection mode in image, is formed corresponding to optical centre and scene point
One ray, how to determine equation of this ray in scene coordinate system is exactly camera calibration problem to be solved.
In calibration process, it is thus necessary to determine that the inner parameter and external parameter of video camera.In three-dimension measuring system, according to projective geometry with
And video camera pin-hole model, the formula for extrapolating transformation relation between scene world coordinate system and pixel coordinate system is as follows.
Wherein (Xwi,Ywi,Zwi, 1) be i-th point of space world coordinates, (ui,vi, 1) and it is i-th point of image coordinate；
mijFor projection matrix M the i-th row jth column element.
, can by by the three-dimensional coordinate of a large amount of known angle point in calibrating block and their pixel coordinates on video camera
To solve projection matrix M.As can be seen from the above equation, the accuracy of detection of angle point directly determines the calculating essence of camera parameters
Degree.
At present, the Corner Detection based on chessboard table images mainly has two ways：A kind of is the method obtained manually, is passed through
Mouse clicks on the position for obtaining angle point one by one；Another is the method for automatic detection, for example, calculated using Harris operators, Susan
Son etc. obtains the corner location of image.The positional information of angle point is obtained manually, and the problem of existing has：Firstly the need of manual intervention.
The speed in camera calibration image processing process is have impact on, and has no idea to realize dynamic calibration；Secondly, the detection of angle point
Precision is unstable, to a certain extent also can be relevant with the experience of operator etc..And in automatic testing method, calculated using Harris
Although son, Susan operators etc. can realize automatic Calibration process, also occur that precision is unstable, situations such as pseudo- angle point, especially
It is that in inclined chessboard table images, and then have impact on the precision of whole camera calibration.
The content of the invention
It is a primary object of the present invention to overcome the shortcoming and deficiency of prior art there is provided one kind based on 180 ° of rotation operators
Demarcation angular-point detection method, have that practical, precision is high, fireballing feature compared to existing angular-point detection method.
In order to achieve the above object, the present invention uses following technical scheme：
A kind of demarcation angular-point detection method based on 180 ° of rotation operators, comprises the steps：
(1) pending uncalibrated image I is obtained：In three-dimension measuring system, calibrating block is positioned over fixed position, adjusted
The brightness of good light source, is shot by video camera and obtains uncalibrated image；
(2) grey scale change degree Z1 is judged：Each point I (i, j) on image is scanned one by one, set up with pixel I (i, j)
Centered on size be n × n square rotation operator template W, calculate operator mould using the method for gray scale mean square deviation is asked for
The flat case of image in plate, i.e. grey scale change degree Z1；
(3) 180 ° of rotation registration Z2 are judged：Judge grey scale change degree Z1 and extract edge and angle point region it
Afterwards, for qualified pixel I (i, j), black and white gray scales in its rotation operator template W are calculated, are carried out in need
Black and white gray scale is exchanged, and is produced new template, is carried out the function of contact ratio calculating again in this new template, i.e., to rotation operator mould
Plate W is carried out after 180 ° of rotations, is obtained rotary template W0, is asked for the average of the absolute value sum of each element in both matrix of differences
Z2；
(4) angle point of image is extracted：After rotation operator template judges grey scale change degree Z1 and rotation registration Z2, obtain
Take out the connected domain of often place angle point region, as interior registration highest point, the angle point of image.
It is preferred that, step (1) is 5 × 5 first to imaging importing template, width is 1 before using rotation operator processing
2-d gaussian filterses device, filter out impurity noise.
It is preferred that, the function g (i, j) of the 2-d gaussian filterses device,WhereinFor template
Centerσ is width, namely smoothness.
It is preferred that, in step (1), the three-dimension measuring system includes projecting apparatus, a camera system and one
Individual calibrating block.
It is preferred that, in step (2), calculate grey scale change degree Z1 mean square deviation function Z1(i, j), be：
In formula,For the average gray of pixel in template W, n is window W width.
It is preferred that, in step (3), judge whether grey scale change degree Z1 meets predetermined threshold value T1, selection threshold value T1 is
0.1, judge Z1Whether (i, j) meets Z1(i,j)>T1, if disclosure satisfy that, is proved centered on current pixel point I (i, j)
Rotation operator template is in marginal point or corner location and non-planar regions, and its grey scale change is larger, continues to calculate its registration
Functional value Z2 processing；If can not meet, without processing, skip.
It is preferred that, in step (3), calculate before 180 ° of rotation registration Z2, judge black and white gray scales, the template gross area
S=n × n, white portion area is S1, selects predetermined threshold value e=(n × n-1)/2, then area is accounted for than α=S1/S,
Wherein W ' (i, j) is the rotation operator template after black and white gray scale is exchanged.
It is preferred that, calculate 180 ° of rotation registration Z2 function Z2(i, j),In formula, n is window W width, and (x, y) is the transfer point in window,
I (i-x, j+y) is that I (i+x, j-y) points rotate the corresponding points after 180 ° centered on I (i, j).
It is preferred that, judge whether rotation registration Z2 meets predetermined threshold value T2, selection threshold value T2 is no more than 0.1, judges Z2
Whether (i, j) meets Z2(i,j)>T2, if disclosure satisfy that, proves the rotation operator template centered on current pixel point I (i, j)
Symmetry enough, registration is higher, pixel I (i, j) be in angle point regional location, can accurately extract its position, instead
Then without processing, skip.
It is preferred that, as the Z obtained by step (2)1Z obtained by (i, j) and step (3)2(i, j), obtains approximate mould residing for angle point
The region of paste, connected domain is obtained out using eight neighborhood, then is superimposed evaluation algorithm and is filtered out angle value Z2 highests are overlapped in connected domain
The angle point of point, as image.
The present invention compared with prior art, has the following advantages that and beneficial effect：
(1) present invention detects angle point by 180 ° of rotation transformations, accelerates detection speed, and process is simple, operability
By force.
(2) present invention has consistency to rotation transformation, and can effectively remove the interference of pseudo- angle point and noise, improves
The degree of accuracy of algorithm and stability.
(3) present invention can be applied not only to general scaling board image, be also applied for the calibrating block image in multiple faces, right
In have certain gradient and distortion uncalibrated image equally have good effect.
(4) present invention can meet needs of expected design, and be calculated with some other automatic detection popular at present
Method is compared, and its applicability is wider, and more preferably, stability is higher for accuracy.
Brief description of the drawings
Fig. 1 is the structural representation of detecting system of the present invention；
Fig. 2 is the flow chart of detection method；
Fig. 3 is the schematic diagram of rotation operator of the present invention；
Fig. 4 is the centrosymmetric schematic diagram of rotation operator of the present invention.
Embodiment
With reference to embodiment and accompanying drawing, the present invention is described in further detail, but embodiments of the present invention are not limited
In this.
As shown in figure 1, the detecting system of the demarcation angular-point detection method based on 180 ° of rotation operators of the present invention, including put
Calibrating block 4 on horizontalization platform 3, projecting apparatus 2 can gather image as light source by video camera 1.In normal calibration process, lead to
The scaling board image more than or equal to 2 is often gathered, and scaling board can not must be placed in same plane.So,
First is cumbersome, and second during scaling board is changed, and can cause certain influence on measurement accuracy unavoidably.Here use
Calibrating block, the cube that constitutes of plate face is demarcated as at least three.As shown in Figure 1.Only need to take pictures using calibrating block
Once, algorithm process is carried out in same figure, thus it is more convenient.
As shown in Fig. 2 being a kind of overall flow figure of angular-point detection method of the invention, its specific operating procedure is as follows：
(1) pending uncalibrated image I is obtained：In with the three-dimension measuring system shown in Fig. 1, calibrating block is positioned over
Fixed position, regulates the brightness of light source, is shot by video camera and obtains uncalibrated image I.Before using rotation operator processing,
It is 5 × 5 first to imaging importing template, width is 1 2-d gaussian filterses device, filters out impurity noise；
(2) grey scale change degree Z1 is judged：Each point I (i, j) on image is scanned one by one, set up with pixel I (i, j)
Centered on size be n × n square rotation operator template W.Operator mould is calculated using the method for gray scale mean square deviation is asked for
The flat case of image in plate, i.e. grey scale change degree Z1.
As shown in figure 3, a kind of principle schematic of the rotation operator designed for the present invention.For image every bit I (i,
J), the template centered on the point is referred to as W.W can be it is circular can also be it is square, here from square template,
If the width of template is n, then the number of pixels in template is n × n.For the accuracy and the accuracy of detection of calculating, W's
Width n should be less than a tessellated length of side.N value is odd number, so ensures that template in each point value of calculating process
All it is integer.
When template W is in region a, image is flat, gray variance value very little, therefore is not marginal point and angle point.
When template W is in region b, its 180 ° of postrotational registrations are higher, are preferable angle points.And c, although image is uneven at d
It is smooth, it may have certain symmetry, but it is relatively low on the symmetry of central point.So, we just can be according to calculating
180 ° of postrotational registrations accurately determine the position where angle point.
Need effectively to reject the flat site in image afterwards, extract marginal point and angle point region.Because flat
In template where the pixel in smooth region, the mean square deviation of pixel grey scale is smaller, so the mean square deviation of design rotation operator
Response rejects flat site, its formula as the gray variance of reflection surrounding pixel gray-value variation severe degree with this
Expression formula, that is, calculate grey scale change degree Z1 mean square deviation function Z1(i, j) is as follows
In formula,For the average gray of pixel in template W, n is window W width.
Judge whether grey scale change degree Z1 meets predetermined threshold value T1.It is about 0.1 to select threshold value T1.Judge Z1(i,j)
Whether Z is met1(i,j)>T1, if disclosure satisfy that, is proved at the rotation operator template centered on current pixel point I (i, j)
In marginal point or corner location and non-planar regions, its grey scale change is larger, continues to calculate at its function of contact ratio value Z2
Reason.If can not meet, without processing, this is non-angle point.
(3) 180 ° of rotation registration Z2 are judged：After judging grey scale change degree Z1, calculate 180 ° of rotation registration Z2 it
Before, for qualified pixel I (i, j), black and white gray scales in its rotation operator template W are calculated, are carried out in need
Black and white gray scale is exchanged, and produces new template.Because, in inclined calibrating block image, directly calculate the registration of the above
Function, the template of same gradient, on the high side or black portions are on the high side for white portion, its result of calculation also has larger
Difference.
Therefore Rule of judgment is added in degree of integration functional procedure is calculated：If in the template of rotation operator, white portion
Area accounts for the more than half of the gross area, then black and white part is carried out into grayvalue transition.Registration letter is carried out again in this new template
Number is calculated, i.e., rotation operator template W is carried out after 180 ° of rotations, obtain rotary template W0, is asked for each in both matrix of differences
The average Z2 of the absolute value sum of element.
Template gross area S=n × n, white portion area is S1, selection predetermined threshold value e=(n × n-1)/2.Then account for area
Than α=S1/S.Situations below is met then to be exchanged.
Wherein W ' (i, j) is the rotation operator template after black and white gray scale is exchanged.
As shown in figure 4, being the centrosymmetric schematic diagram of rotation operator of the present invention.For some pixel A (i+ in template
X, j-y) on central point I (i, j), symmetrically point is A1 (i+x, j-y).The registration response for defining rotation operator is template W
In on symmetrical every a pair of the pixels of central point I (i, j) gray scale difference of two squares absolute value average value, then the function of contact ratio table
Show formula Z2(i, j) is as follows：
In formula, n is window W width, and (x, y) is the transfer point in window.I (i-x, j+y) is I (i+x, j-y) points with I
The corresponding points after 180 ° are rotated centered on (i, j).The function of contact ratio calculated value Z2 of feature angle point on uncalibrated image is smaller；And
Boundary point and noise etc., due to asymmetry, one side grey scale pixel value is big, and another side grey scale pixel value is small, so its Z2 is larger.
Certain point I (i, j) the function of contact ratio calculated value Z2 is on image, pixel in the range of the wicket centered on the pixel
The reflection of the spatial symmetry of intensity profile.
Judge whether rotation registration Z2 meets predetermined threshold value T2.Threshold value T2 is selected to be no more than 0.1.Judge Z2(i, j) is
It is no to meet Z2(i,j)>T2, if disclosure satisfy that, proves the symmetrical of rotation operator template centered on current pixel point I (i, j)
Property enough, registration is higher, pixel I (i, j) be in angle point regional location, can accurately extract its position.On the contrary then nothing
Processing is needed, belongs to non-angle point.
(4) angle point of image is extracted：By the schematic diagram of Fig. 4 rotation operator, it is seen that this is in being with pixel I (i, j)
Heart rotation operator template W (n × n square matrices), is carried out after 180 ° of rotations, obtained new rotary template W0, both
Subtract each other again, ask for the average of the absolute value sum of each element in matrix of differences, i.e. the function of contact ratio value Z2.
After rotation operator template judges grey scale change degree Z1 and rotation registration Z2, every place's angle point location is obtained out
After the connected domain in domain, then the angle point in each region is extracted respectively.Next eight neighborhood calculating is carried out to whole image, obtained
Go out the connected domain of each angle point region.Interior registration highest point, the as angle point of image.If the company extracted in previous step
Pixel number in logical domain is l, works as l<2 or l>When 8, the pixel in this connected domain is noise either marginal point, is gone
Remove；If 1<l<9, then filter out and angle value Z2 highest points, the as angle point of image are overlapped in connected domain.So far, uncalibrated image
Corner Detection is completed.
Above-described embodiment is preferably embodiment, but embodiments of the present invention are not by above-described embodiment of the invention
Limitation, other any Spirit Essences without departing from the present invention and the change made under principle, modification, replacement, combine, simplification,
Equivalent substitute mode is should be, is included within protection scope of the present invention.
Claims (10)
1. a kind of demarcation angular-point detection method based on 180 ° of rotation operators, it is characterised in that comprise the steps：
(1) pending uncalibrated image I is obtained：In three-dimension measuring system, calibrating block is positioned over fixed position, light is regulated
The brightness in source, is shot by video camera and obtains uncalibrated image；
(2) grey scale change degree Z1 is judged：Each point I (i, j) on image is scanned one by one, during foundation is with pixel I (i, j)
The size of the heart is n × n square rotation operator template W, and rotation operator mould is calculated using the method for gray scale mean square deviation is asked for
The flat case of image in plate, i.e. grey scale change degree Z1；
(3) 180 ° of rotation registration Z2 are judged：It is right after judging grey scale change degree Z1 and extracting edge and angle point region
In qualified pixel I (i, j), black and white gray scales in its rotation operator template W are calculated, black-white-gray in need is carried out
Degree exchange, produces new template, carries out the function of contact ratio calculating again in this new template, i.e., rotation operator template W is entered
After 180 ° of rotations of row, new rotation operator template W0 is obtained, the absolute value sum of each element in both matrix of differences is asked for
Average Z2；
(4) angle point of image is extracted：After rotation operator template W judges grey scale change degree Z1 and rotation registration Z2, obtain
Go out registration highest point, the as angle point of image in the connected domain of every place's angle point region.
2. the demarcation angular-point detection method according to claim 1 based on 180 ° of rotation operators, it is characterised in that step
(1) it is 5 × 5 first to imaging importing template before using rotation operator processing, width is 1 2-d gaussian filterses device, filter
Removal of impurity noise.
3. the demarcation angular-point detection method according to claim 2 based on 180 ° of rotation operators, it is characterised in that described two
The function g (i, j) of Gaussian filter is tieed up,WhereinFor template centerσ is width
Degree, namely smoothness.
4. the demarcation angular-point detection method according to claim 1 based on 180 ° of rotation operators, it is characterised in that step
(1) in, the three-dimension measuring system includes a projecting apparatus, a camera system and a calibrating block.
5. the demarcation angular-point detection method according to claim 1 based on 180 ° of rotation operators, it is characterised in that step
(2) in, grey scale change degree Z1 mean square deviation function Z is calculated1(i, j), be：
<mrow>
<msub>
<mi>Z</mi>
<mn>1</mn>
</msub>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mrow>
<msub>
<mo>&Sigma;</mo>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>)</mo>
<mo>&Element;</mo>
<mi>W</mi>
</mrow>
</msub>
<msup>
<mrow>
<mo>(</mo>
<mi>I</mi>
<mo>(</mo>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
<mo>)</mo>
<mo>-</mo>
<mover>
<mi>I</mi>
<mo>&OverBar;</mo>
</mover>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
</mrow>
<mrow>
<mi>n</mi>
<mo>&times;</mo>
<mi>n</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
</mfrac>
</mrow>
In formula,For the average gray of pixel in rotation operator template W, n is rotation operator template W width.
6. the demarcation angular-point detection method according to claim 5 based on 180 ° of rotation operators, it is characterised in that step
(3) in, judge whether grey scale change degree Z1 meets predetermined threshold value T1, selection threshold value T1 is 0.1, judges Z1Whether (i, j) meets Z1
(i,j)>T1, if disclosure satisfy that, prove rotation operator template centered on current pixel point I (i, j) be in marginal point or
Corner location and non-planar regions, its grey scale change are larger, continue to calculate its and rotate registration Z2 processing；If can not expire
Foot, without processing, is skipped.
7. the demarcation angular-point detection method according to claim 1 based on 180 ° of rotation operators, it is characterised in that step
(3) in, calculate before 180 ° of rotation registration Z2, judge black and white gray scales, template gross area S=n × n, white portion face
Product is S1, selection predetermined threshold value e=(n × n-1)/2, then accounts for area than α=S1/S,
<mfenced open = "{" close = "">
<mtable>
<mtr>
<mtd>
<mrow>
<mi>&alpha;</mi>
<mo>&le;</mo>
<mi>e</mi>
<mo>,</mo>
</mrow>
</mtd>
<mtd>
<mrow>
<msup>
<mi>W</mi>
<mo>&prime;</mo>
</msup>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mi>W</mi>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>)</mo>
</mrow>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<mi>&alpha;</mi>
<mo>></mo>
<mi>e</mi>
<mo>,</mo>
</mrow>
</mtd>
<mtd>
<mrow>
<msup>
<mi>W</mi>
<mo>&prime;</mo>
</msup>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mn>1</mn>
<mo>-</mo>
<mi>W</mi>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>)</mo>
</mrow>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
Wherein W ' (i, j) is the rotation operator template after black and white gray scale is exchanged.
8. the demarcation angular-point detection method according to claim 5 based on 180 ° of rotation operators, it is characterised in that calculate
180 ° of rotation registration Z2 function Z2(i, j),In formula, n is rotation
Operator template W width, (x, y) is the transfer point in template, and I (i-x, j+y) is I (i+x, j-y) points centered on I (i, j)
Corresponding points after 180 ° of rotation.
9. the demarcation angular-point detection method according to claim 8 based on 180 ° of rotation operators, it is characterised in that judge rotation
Turn whether registration Z2 meets predetermined threshold value T2, selection threshold value T2 is no more than 0.1, judges Z2Whether (i, j) meets Z2(i,j)>
T2, if disclosure satisfy that, proves the symmetry of rotation operator template W centered on current pixel point I (i, j) enough, registration
Sufficiently high, pixel I (i, j) is in angle point regional location, can accurately extract its position, it is on the contrary then without handle, skip.
10. the demarcation angular-point detection method according to claim 8 based on 180 ° of rotation operators, it is characterised in that by step
Suddenly the Z obtained by (2)1Z obtained by (i, j) and step (3)2(i, j), obtains region approximately fuzzy residing for angle point, utilizes eight neighborhood
Connected domain is obtained out, then is superimposed evaluation algorithm and filters out rotation registration Z2 highest points, the as angle point of image in connected domain.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510127733.6A CN104751458B (en) | 2015-03-23 | 2015-03-23 | A kind of demarcation angular-point detection method based on 180 ° of rotation operators |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510127733.6A CN104751458B (en) | 2015-03-23 | 2015-03-23 | A kind of demarcation angular-point detection method based on 180 ° of rotation operators |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104751458A CN104751458A (en) | 2015-07-01 |
| CN104751458B true CN104751458B (en) | 2017-08-25 |
Family
ID=53591077
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510127733.6A Active CN104751458B (en) | 2015-03-23 | 2015-03-23 | A kind of demarcation angular-point detection method based on 180 ° of rotation operators |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104751458B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105551048A (en) * | 2015-12-21 | 2016-05-04 | 华南理工大学 | Space surface patch-based three-dimensional corner detection method |
| CN106023171B (en) * | 2016-05-12 | 2019-05-14 | 惠州学院 | A kind of image angular-point detection method based on turning radius |
| CN106780611A (en) * | 2016-12-10 | 2017-05-31 | 广东文讯科技有限公司 | One kind uses intelligent terminal camera angular-point detection method |
| CN109658372B (en) * | 2017-10-10 | 2021-01-26 | 凌云光技术股份有限公司 | Image uniformity evaluation method and device |
| CN107941241B (en) * | 2017-10-10 | 2021-04-06 | 天津大学 | Resolution board for aerial photogrammetry quality evaluation and use method thereof |
| CN114140521A (en) * | 2020-09-04 | 2022-03-04 | 深圳光峰科技股份有限公司 | Method, device and system for identifying projection position and storage medium |
| CN113376170B (en) * | 2021-06-16 | 2023-01-10 | 博众精工科技股份有限公司 | Calibration method and calibration block of product appearance detection equipment |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101763632A (en) * | 2008-12-26 | 2010-06-30 | 华为技术有限公司 | Method for demarcating camera and device thereof |
| CN101980292A (en) * | 2010-01-25 | 2011-02-23 | 北京工业大学 | Regular octagonal template-based board camera intrinsic parameter calibration method |
| CN103198319A (en) * | 2013-04-11 | 2013-07-10 | 武汉大学 | Method of extraction of corner of blurred image in mine shaft environment |
| WO2013182080A1 (en) * | 2012-06-08 | 2013-12-12 | 华为技术有限公司 | Parameter calibration method and device |
| CN103679720A (en) * | 2013-12-09 | 2014-03-26 | 北京理工大学 | Fast image registration method based on wavelet decomposition and Harris corner detection |
| CN103927750A (en) * | 2014-04-18 | 2014-07-16 | 上海理工大学 | Detection method of checkboard grid image angular point sub pixel |
| CN104331900A (en) * | 2014-11-25 | 2015-02-04 | 湖南科技大学 | Corner sub-pixel positioning method in CCD (charge coupled device) camera calibration |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8855442B2 (en) * | 2012-04-30 | 2014-10-07 | Yuri Owechko | Image registration of multimodal data using 3D-GeoArcs |
-
2015
- 2015-03-23 CN CN201510127733.6A patent/CN104751458B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101763632A (en) * | 2008-12-26 | 2010-06-30 | 华为技术有限公司 | Method for demarcating camera and device thereof |
| CN101980292A (en) * | 2010-01-25 | 2011-02-23 | 北京工业大学 | Regular octagonal template-based board camera intrinsic parameter calibration method |
| WO2013182080A1 (en) * | 2012-06-08 | 2013-12-12 | 华为技术有限公司 | Parameter calibration method and device |
| CN103198319A (en) * | 2013-04-11 | 2013-07-10 | 武汉大学 | Method of extraction of corner of blurred image in mine shaft environment |
| CN103679720A (en) * | 2013-12-09 | 2014-03-26 | 北京理工大学 | Fast image registration method based on wavelet decomposition and Harris corner detection |
| CN103927750A (en) * | 2014-04-18 | 2014-07-16 | 上海理工大学 | Detection method of checkboard grid image angular point sub pixel |
| CN104331900A (en) * | 2014-11-25 | 2015-02-04 | 湖南科技大学 | Corner sub-pixel positioning method in CCD (charge coupled device) camera calibration |
Non-Patent Citations (3)
| Title |
|---|
| "使用12 像素对称模板的棋盘格内角点检测";戴士杰等;《红外与激光工程》;20140430;第43卷(第4期);1306-1311 * |
| "基于双旋转模板的黑白棋盘角点检测算法";凌江;《图形、图像与多媒体》;20141231;第33卷(第21期);42-45 * |
| "用于摄像机标定的棋盘图像角点检测新算法";杨幸芳等;《仪器仪表学报》;20110531;第32卷(第5期);1109-1113 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104751458A (en) | 2015-07-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104751458B (en) | A kind of demarcation angular-point detection method based on 180 ° of rotation operators | |
| CN103345755B (en) | A kind of Chessboard angular point sub-pixel extraction based on Harris operator | |
| CN107993263B (en) | Automatic calibration method for panoramic system, automobile, calibration device and storage medium | |
| CN106340044B (en) | Join automatic calibration method and caliberating device outside video camera | |
| CN103530880B (en) | Based on the camera marking method of projection Gaussian network pattern | |
| CN104048815B (en) | Method and system for measuring distortion of lens | |
| CN107607040A (en) | A kind of three-dimensional scanning measurement device and method suitable for High Reflective Surface | |
| CN109859226A (en) | A kind of detection method of the X-comers sub-pix of figure segmentation | |
| CN103824275B (en) | Saddle dots structure and the system and method for determining its information are searched in the picture | |
| CN111080662A (en) | Lane line extraction method and device and computer equipment | |
| CN108305286A (en) | Multi-view stereo vision foot type method for three-dimensional measurement, system and medium based on color coding | |
| CN106600647A (en) | Binocular visual multi-line projection structured light calibration method | |
| CN106937109B (en) | The method that low cost judges resolution ratio of camera head level | |
| CN111861980B (en) | Imaging detection method, electronic equipment and computer readable storage medium | |
| CN107869954A (en) | A kind of binocular vision volume weight measuring system and its implementation | |
| JP2013174547A (en) | Stereo three-dimensional measuring instrument | |
| CN104574312A (en) | Method and device of calculating center of circle for target image | |
| CN107490342A (en) | A kind of cell phone appearance detection method based on single binocular vision | |
| CN110866882A (en) | Layered joint bilateral filtering depth map restoration algorithm based on depth confidence | |
| CN107356213A (en) | Optical filter concentricity measuring method and terminal device | |
| CN106952262A (en) | A kind of deck of boat analysis of Machining method based on stereoscopic vision | |
| CN106485678A (en) | A kind of depth of field space-time consistency based on spatio-temporal filtering and the enhanced method of precision | |
| CN104537627B (en) | A kind of post-processing approach of depth image | |
| CN108986721A (en) | A kind of test pattern generation method for display panel detection | |
| CN110378967B (en) | Virtual target calibration method combining grating projection and stereoscopic vision |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |