CN103177439B - A kind of automatic calibration method based on black and white lattice corners Matching - Google Patents
A kind of automatic calibration method based on black and white lattice corners Matching Download PDFInfo
- Publication number
- CN103177439B CN103177439B CN201210486358.0A CN201210486358A CN103177439B CN 103177439 B CN103177439 B CN 103177439B CN 201210486358 A CN201210486358 A CN 201210486358A CN 103177439 B CN103177439 B CN 103177439B
- Authority
- CN
- China
- Prior art keywords
- point
- image
- coordinate
- angle point
- black
- 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.)
- Expired - Fee Related
Links
Landscapes
- Image Processing (AREA)
Abstract
The invention discloses a kind of automatic calibration method based on black and white lattice corners Matching, adopt the demarcation cloth of black and white lattice as calibrating template and improvement harris Corner Detection Algorithm respectively detected image angle point and the position relationship set up between angle point forms Corner-point Grids; Distortion correction and splicing are carried out to picture; And the homography matrix of angle point true coordinate in the angular coordinate calculated in orthoscopic image and final virtual panoramic aerial view; The position getting a bird's eye view video camera according to virtual and homography matrix carry out view transformation, obtain the corresponding relation between picture point in this coordinate system and original graph picture point, generate look-up table.The present invention is directed to improvement Harris Corner Detection Algorithm simply practical, the judgement to angle point position relationship can well be realized simultaneously, form correct Corner-point Grids; Export after adopting the mode of tabling look-up to realize demarcating and look map generalization, speed soon, well meets the requirement of real-time.
Description
Technical field
The present invention relates to Digital Image Processing and computer vision technique, especially, relate to the camera marking method of panoramic parking system.
Background technology
China is the serious country of a traffic hazard, and annual death toll and fatal rate all allow of no optimist.Reduce traffic hazard and not only in subjective enhancing awareness of safety, objectively more will will improve the reliability of the safe driving of automobile own.Existing miscellaneous auxiliary driving instrument is exactly to address this problem.Automotive safety safeguards system substantially can be divided into passive and active system, and the former mainly comprises securing band, air bag etc., although accident casualty degree can be reduced, and can not the generation of Accident prevention.Latter mainly utilize various kinds of sensors as, ultrasound wave, radar, infrared heat sensor and video camera etc.They can provide the traffic informations such as barrier for driver's decision-making.Meanwhile, active safety safeguards system also constitutes the important component part of intelligent transportation system.In above-mentioned all methods, video camera has the advantages such as low cost, easy care and high integration, is therefore widely used.
Along with the fast development of image procossing and computer vision, increasing advanced technology is applied to vehicle electric field.Traditional auxiliary driving based on image, only installs backup camera at automobile tail, can only cover the region that motor vehicle environment is limited, and the vision dead zone in car both sides and front adds the hidden danger of safe driving undoubtedly.In order to expand the field range of driver, just necessary perception automobile surrounding 360 ° of environment, this just needs information fusion and registration between multiple vision sensor.Mutual alignment between multiple vision sensor and attitude Relation Parameters obtain from specific survey sensor, and this depends critically upon accurate Position and orientation parameters with regard to making the result of fusion registration.
Summary of the invention
The present invention is directed to the panoramic shooting backup system of automobile, the invention provides a kind of virtual design proposal getting a bird's eye view panorama sketch that can obtain around vehicle body.
Based on an automatic calibration method for black and white lattice corners Matching, comprise the following steps:
(1) adopt the demarcation cloth of black and white lattice as calibrating template, the checkerboard image of the four direction all around of the automobile that camera acquisition arrives is adopted to the angle point of the harris Corner Detection Algorithm difference detected image improved;
(2) the inner angle point of black and white lattice is calculated and the position relationship set up between angle point by black and white lattice matching algorithm to the four width images detected in (1), form Corner-point Grids;
(3) the center of distortion point of four fault images is calculated according to the Corner-point Grids obtained in (2);
(4) carrying out the distortion correction of image according to the center of distortion point of the fault image obtained in (3), focal length and camera lens model, is distortionless image four width image rectifications;
(5) calculate the final size exporting view according to the size in vehicle commander, overall width and the actual visual field, and the angle point calculating four width orthoscopic images according to the final size exporting view virtual get a bird's eye view relative coordinate in panorama sketch and virtual get a bird's eye view splicing in panorama sketch view and merge after true coordinate position;
(6) according to the distortionless image obtained in (4), the virtual homography matrix getting a bird's eye view angle point true coordinate in panorama sketch obtained in angular coordinate in orthoscopic image and (5) is obtained;
(7) carry out view transformation according to the virtual position getting a bird's eye view video camera with the homography matrix in (6), four width image conversions are got a bird's eye view the virtual coordinate system of getting a bird's eye view video camera of video camera as initial point to using virtual;
(8) according to the virtual coordinate system getting a bird's eye view video camera in (7), obtain the corresponding relation between picture point in this coordinate system and original graph picture point, generate look-up table.
Described improvement harris Corner Detection innovatory algorithm is:
(1) horizontal gradient operator is adopted
，
VG (vertical gradient) operator
，
Try to achieve horizontal gradient and the VG (vertical gradient) of each pixel
with
, and generator matrix
, wherein
.
(2) right
,
and
respectively by second order Gauss filter function
Carry out gaussian filtering, wherein
for
,
and
in the standard deviation of element,
with
it is the difference of the coordinate of the element in Gaussian template and the coordinate of template center's element.Right respectively
,
with
after three matrixes carry out gaussian filtering, the new matrix of recomposition
.
(3) interest value of each pixel on original image is calculated, i.e. R value:
，
The matrix that each pixel is corresponding
two corresponding eigenwerts.
(4) set up one with the full null matrix result of the equal size of image to preserve angle point result, search for the matrix that all-pair is answered successively
eigenwert, see and whether be both greater than certain threshold value, the point satisfied condition assignment in result is 1.
(5) set the window of a specific size, allow window move on image, may there is the angle point obtained in multiple previous step in a window, select the point wherein making the value of R minimum as angle point, all the other assignment are 0.
Described black and white lattice matching algorithm is:
(1) carry out gaussian filtering to image, wherein Gaussian template is:
；
(2) according to coordinate angle point being detected, a default numerical value is added at the coordinate figure in x direction and y direction respectively at each angular coordinate, if the coordinate figure in coordinate figure x direction or y direction is less than 0 or be greater than the maximum coordinates of image after adding default numerical value, then represent that this angle point is image border angle point, should be cast out.
(3) or edge angle point inner with black and white lattice is for the center of circle, with the half of the spacing of adjacent two angle points for radius does circle respectively, around circle one week, if it is not twice or four times that the color of circumference changes, is then removed by the angle point of correspondence.
(4) a few step such as position relationship, angular relationship, distance relation, black and white lattice rule by judging between angle point judges the relative position found out between angle point.
(5) by the judgement of position proximity relations between all angle points, Corner-point Grids is formed.
Described black and white lattice rule is: for angle point 1 and angle point 2, get the connecting line of angle point 1 and angle point 2 for long, and the half of getting this connecting line is widely do rectangle in the both sides of connecting line respectively, and be that two rectangles make diagonal line respectively, if the diagonal line of two rectangles is consistent respectively with the color of length corresponding except connecting line, then angle point 1 is adjacent with angle point 2, otherwise is non-conterminous.
The center of distortion point of described calculating four fault images is calculated by the mode of fitting circle, and method is as follows:
(1) often going and often arranging respectively by least square fitting circle Corner-point Grids.
(2) circle of two row matchings and the circle of two row matchings is got successively to ask fish-eye center of distortion point.
(3) all center of distortions point obtained is averaged as final center of distortion point.
The method of described distortion correction is:
Wherein
it is image to be corrected
on a bit,
it is a image after calibration
corresponding a bit,
a little
arrive
the Euclidean distance of picture centre.
be
photocentre,
be the focal length of camera, camera lens projection model is
,
a little
incident angle.
be
central point.
The method for solving of described homography matrix is:
Set up and virtually get a bird's eye view video camera angular coordinate
with the angular coordinate of orthoscopic image
homography matrix L:
Set up and virtually get a bird's eye view video camera angular coordinate
with the angular coordinate of orthoscopic image
homography matrix L formula:
，
；
When the region taken is surface level,
, then
，
；
Wherein
be the element in L, can calculate according to the Corner-point Grids coordinate of the image after four width distortion corrections and the virtual position relationship getting a bird's eye view the coordinate of panorama sketch.
The method of described view transformation is:
，
The point in the angular coordinate system of orthoscopic image is calculated according to above formula
the virtual level coordinates system getting a bird's eye view video camera angle point after carrying out visual transform
.
The generation method of described lookup table is, the virtual each point getting a bird's eye view panorama sketch can calculate the coordinate of the point of changing the time after distortion correction according to homography matrix, calculate this coordinate in original fault image through the inverse transformation of distortion correction again, and relation generates lookup table according to this.
In sum, the present invention has following beneficial effect: (1) the present invention can reach good angle point grid effect, for the deficiency of Harris Corner Detection Algorithm, proposes a kind of simply innovatory algorithm of practicality; (2) adopt the algorithm finding angle point relation, the judgement to angle point position relationship can well be realized, form correct Corner-point Grids; (3) export after the present invention adopts the mode of tabling look-up to realize demarcating and look map generalization, speed is fast, well meets the requirement of real-time.
Accompanying drawing explanation
Fig. 1 is the overall flow schematic diagram of the automatic calibration method based on black and white lattice corners Matching of the present invention;
Fig. 2 is the black and white case marker solid plate schematic diagram demarcated for coordinate of the present invention;
Fig. 3 is the fish eye images of fish eye lens shooting;
Fig. 4 is the image after distortion correction;
Fig. 5 is that black and white lattice edge angle point judges schematic diagram;
Fig. 6 is black and white lattice rule schematic diagram.
Embodiment
In order to allow those skilled in the art can understand technical scheme of the present invention better, below in conjunction with accompanying drawing, the invention will be further elaborated.
As shown in Figure 1, present invention is disclosed a kind of automatic calibration method based on black and white lattice corners Matching, comprise the following steps: (1) adopts the demarcation cloth of black and white lattice as calibrating template (as shown in Figure 2), carry out black and white lattice Corner Detection to the checkerboard image of the four direction all around of the automobile collected, Corner Detection uses the harris Corner Detection Algorithm improved.
(2) the inner angle point of black and white lattice is calculated and the position relationship set up between angle point by black and white lattice matching algorithm to the four width images detected in (1), form Corner-point Grids.
(3) the center of distortion point of four fault images is calculated according to the Corner-point Grids obtained in (2).
(4) carrying out the distortion correction of image according to the center of distortion point of the fault image obtained in (3), focal length and camera lens model, is distortionless image four width image rectifications.
(5) calculate the final size (namely following virtual get a bird's eye view panorama sketch) exporting view according to the size in vehicle commander, overall width and the actual visual field, and the angle point calculating four width images according to the final size exporting view gets a bird's eye view relative coordinate in panorama sketch and in the virtual true coordinate position got a bird's eye view in panorama sketch view after splicing and fusion virtual.
(6) according to the distortionless image obtained in (4), the virtual homography matrix getting a bird's eye view angle point true coordinate in panorama sketch obtained in angular coordinate in orthoscopic image and (5) is obtained.
(7) carry out view transformation according to the virtual position getting a bird's eye view video camera with the homography matrix in (6), four width image conversions are got a bird's eye view the virtual coordinate system of getting a bird's eye view video camera of video camera as initial point to using virtual;
(8) according to the virtual coordinate system getting a bird's eye view video camera in (7), obtain the corresponding relation between picture point in this coordinate system and original graph picture point, generate look-up table.
The fish eye images of flake video camera shooting as shown in Figure 3, uses harris Corner Detection Algorithm to process fish eye images, first adopts horizontal gradient operator
，
VG (vertical gradient) operator
，
Try to achieve horizontal gradient and the VG (vertical gradient) of each pixel
with
, and generator matrix:
, wherein
.
.
Then, right
,
and
respectively by second order Gauss filter function
，
Carry out gaussian filtering, wherein
for the standard deviation of data,
with
it is the difference of the coordinate of the element in template and the coordinate of template center's element.Right respectively
,
with
after three matrixes carry out gaussian filtering, the new matrix of recomposition
.
If
with
the matrix that certain pixel concrete is corresponding
two eigenwerts, from KLT Corner Detection Algorithm, when
with
when being all greater than certain threshold value, just can determine that this point is angle point.Set up one with the full null matrix result of the equal size of image to preserve angle point result, search for the matrix that all-pair is answered successively
eigenwert, see and whether be both greater than certain threshold value, the point satisfied condition assignment in result is 1.
But due to the tessellated singularity of black and white, just only there is angle point in the point of intersection of black and white lattice, so unnecessary angle point or pseudo-angle point must be removed, in order to obtain real angle point, first calculate the interest value of each pixel on original image, i.e. R value:
，
The window of a setting specific size, window is allowed to move on image, the angle point obtained in multiple previous step may be there is in a window, select the point wherein making the value of R minimum as angle point, all the other assignment are 0, then now result matrix intermediate value be 1 pixel corresponding to element be real angle point.
After Corner Detection completes, corners Matching is carried out to the angle point detected, and forms Corner-point Grids.
First carry out gaussian filtering to image, make its obfuscation, object makes the border of black and white lattice intersection in digital picture so unobvious, is conducive to subsequent treatment.Gaussian template is:
。
Then according to the coordinate of angle point detected, add 10 in x direction and y direction respectively at each angular coordinate, if x direction or y direction are less than 0 or be greater than the maximum coordinates of image, then represent that this angle point is image border angle point, should be cast out.As shown in Figure 5, with black and white lattice, inner or edge angle point is for the center of circle, with the half of the spacing of two angle points for radius does circle, around this circle one week, then the angle point of black and white lattice inside can experience four color changes, and the angle point at black and white lattice edge can experience twice color change.Judge whether it belongs to black and white lattice inside or edge angle point, if do not belong to, is got rid of by this angle point.
The relative position found out between angle point is judged by a few step such as position relationship, angular relationship, distance relation, black and white lattice rule judging between angle point.Wherein, as shown in Figure 6, black and white lattice rule is: for angle point 1 and angle point 2, get the connecting line of angle point 1 and angle point 2 for long, and the half of getting this connecting line is widely do rectangle in the both sides of connecting line respectively, and be that two rectangles make diagonal line respectively, if the diagonal line of two rectangles is consistent respectively with the color of length corresponding except connecting line, then angle point 1 is adjacent with angle point 2, otherwise is non-conterminous.
By the judgement of position proximity relations between all angle points, form Corner-point Grids, after forming Corner-point Grids, then asked the center of distortion point of image by fitting circle, concrete steps are as follows:
Often going and often arranging respectively by least square fitting circle Corner-point Grids, then get the circle of two row matchings and the circle of two row matchings successively to ask fish-eye center of distortion point, and get all center of distortions point and average as final center of distortion point.
Now, according to center of distortion point, distortion correction is carried out to fault image.Image before supposing distortion correction
arbitrary coordinate be
, image after correction
point
more corresponding coordinate is
, the coordinate of center of distortion point is
(namely
photocentre).The process of distortion correction is exactly find out the relation between two coordinate systems, the process of image after then obtaining distortion correction by bilinear interpolation.Its method is as follows:
，
Wherein,
a little
arrive
the Euclidean distance of picture centre.
be the focal length of camera, camera lens projection model is
,
a little
incident angle.
be
central point.
As shown in above formula, first coordinates computed
to the distance of fault image central point
, pass through
obtain incident angle
.
By camera lens model obtain in fault image with
corresponding point
to the distance of fault image central point
.By obtaining above
with
corresponding relation
Thus the image after can distortion correction being obtained, as shown in Figure 4 (this figure is taken by material object and obtains, and has a little deformation, but does not affect expression).
Generate and four width figure are fused into one and virtually get a bird's eye view panorama sketch, just must calculate according to the size in vehicle commander, overall width and the actual visual field size finally exporting view (namely virtual get a bird's eye view panorama sketch), and the angle point calculating four width images thus virtual get a bird's eye view relative coordinate in panorama sketch view and virtual get a bird's eye view splicing in panorama sketch view and merge after true coordinate position.
Virtual panorama sketch of getting a bird's eye view according to generating sets up the virtual coordinate system getting a bird's eye view video camera
, then with distortion correction after image
(coordinate is herein
), according to image after the coordinate of the virtual angle point got a bird's eye view in panorama sketch and distortion correction angular coordinate between relation can obtain distortion correction after image and the virtual homography matrix got a bird's eye view between panorama sketch, have according to DLT algorithm:
，
，
When the region taken is surface level, can think
, then:
，
，
Wherein
it is the element in L, eight yuan of linear function groups can be obtained according to the Corner-point Grids coordinate of the image after four width distortion corrections and the virtual position relationship getting a bird's eye view the coordinate of panorama sketch, re-use the optimum solution that least square method solves this system of equations, solve homography matrix L.
The method of described view transformation is:
，
The point in the angular coordinate system of orthoscopic image is calculated according to above formula
the virtual level coordinates system getting a bird's eye view video camera angle point after carrying out visual transform
.
Virtual certain point getting a bird's eye view panorama sketch
can be calculated a little by homography matrix
the coordinate of the point after distortion correction
, then calculate a little through the inverse transformation of distortion correction
coordinate points in original fault image
.The lookup table of the pixel corresponding relation of panorama sketch and original fault image can be got a bird's eye view by generating virtual according to said method, this completes final system calibrating, just corresponding relation between the virtual image pixel getting a bird's eye view panorama sketch and original fault image can be directly found out according to lookup table when carrying out the display of virtual panorama general view, then generate view by bilinear interpolation, so just can form one and complete virtually get a bird's eye view panorama sketch.
The present embodiment is more excellent embodiment of the present invention; it should be noted that; when not deviating from the present invention's spirit and essence thereof; those of ordinary skill in the art are when making various corresponding change and distortion according to the present invention, but these change and are out of shape the protection domain that all should belong to the claim appended by the present invention.
Claims (8)
1. based on an automatic calibration method for black and white lattice corners Matching, it is characterized in that: comprise the following steps:
(1) adopt the demarcation cloth of black and white lattice as calibrating template, the checkerboard image of the four direction all around of the automobile that camera acquisition arrives is adopted to the angle point of the harris Corner Detection Algorithm difference detected image improved;
(2) the inner angle point of black and white lattice is calculated and the position relationship set up between angle point by black and white lattice matching algorithm to the four width images detected in (1), form Corner-point Grids;
(3) the center of distortion point of four fault images is calculated according to the Corner-point Grids obtained in (2);
(4) carrying out the distortion correction of image according to the center of distortion point of the fault image obtained in (3), focal length and camera lens model, is distortionless image four width image rectifications;
(5) calculate the final size exporting view according to the size in vehicle commander, overall width and the actual visual field, and the angle point calculating four distortionless images according to the final size exporting view virtual get a bird's eye view relative coordinate in panorama sketch and virtual get a bird's eye view splicing in panorama sketch view and merge after true coordinate position;
(6) according to the distortionless image obtained in (4), the virtual homography matrix getting a bird's eye view angle point true coordinate in panorama sketch obtained in angular coordinate in orthoscopic image and (5) is obtained;
(7) carry out view transformation according to the virtual position getting a bird's eye view video camera with the homography matrix in (6), four distortionless image conversions are got a bird's eye view the virtual coordinate system of getting a bird's eye view video camera of video camera as initial point to using virtual;
(8) according to the virtual coordinate system getting a bird's eye view video camera in (7), obtain the corresponding relation between picture point in this coordinate system and original graph picture point, generate look-up table;
The method of described distortion correction is:
Wherein
it is image to be corrected
on a bit,
it is a image after calibration
corresponding a bit,
a little
arrive
the Euclidean distance of picture centre;
be
photocentre,
be the focal length of camera, camera lens projection model is
,
a little
incident angle;
be
central point.
2. a kind of automatic calibration method based on black and white lattice corners Matching according to claim 1, is characterized in that, the harris Corner Detection Algorithm of described improvement is:
(1) horizontal gradient operator is adopted
，
VG (vertical gradient) operator
，
Try to achieve horizontal gradient and the VG (vertical gradient) of each pixel
with
, and generator matrix
, wherein
.
(2) right
,
and
respectively by second order Gauss filter function
Carry out gaussian filtering, wherein
for
,
and
in the standard deviation of element,
with
it is the difference of the coordinate of the element in Gaussian template and the coordinate of template center's element; Right respectively
,
with
after three matrixes carry out gaussian filtering, the new matrix of recomposition
;
(3) interest value of each pixel on original image is calculated, i.e. R value:
，
Wherein
with
it is the matrix that each pixel is corresponding
two corresponding eigenwerts;
(4) set up one with the full null matrix result of the equal size of image to preserve angle point result, search for the matrix that all pixels are corresponding successively
eigenwert, see and whether be both greater than certain threshold value, the point satisfied condition assignment in result is 1;
(5) set the window of a specific size, allow window move on image, may there is the angle point obtained in multiple previous step in a window, select the point wherein making the value of R minimum as angle point, all the other assignment are 0.
3. a kind of automatic calibration method based on black and white lattice corners Matching according to claim 1, is characterized in that, described black and white lattice matching algorithm is
(1) carry out gaussian filtering to image, wherein Gaussian template is:
；
(2) according to the coordinate of the angle point detected, a default numerical value is added at the coordinate figure in x direction and y direction respectively at each angular coordinate, if the coordinate figure in coordinate figure x direction or y direction is less than 0 or be greater than the maximum coordinates of image after adding default numerical value, then represent that this angle point is image border angle point, should be cast out;
(3) or edge angle point inner with black and white lattice is for the center of circle, with the half of the spacing of adjacent two angle points for radius does circle respectively, around circle one week, if it is not twice or four times that the color of circumference changes, is then removed by the angle point of correspondence;
(4) a few step such as position relationship, angular relationship, distance relation, black and white lattice rule by judging between angle point judges the relative position found out between angle point;
(5) by the judgement of position proximity relations between all angle points, Corner-point Grids is formed.
4. a kind of automatic calibration method based on black and white lattice corners Matching according to claim 3, it is characterized in that, described black and white lattice rule is: for angle point 1 and angle point 2, get the connecting line of angle point 1 and angle point 2 for long, and the half of getting this connecting line is widely do rectangle in the both sides of connecting line respectively, and be that two rectangles make diagonal line respectively, if the diagonal line of two rectangles is consistent respectively with the color of length corresponding except connecting line, then angle point 1 is adjacent with angle point 2, otherwise is non-conterminous.
5. a kind of automatic calibration method based on black and white lattice corners Matching according to claim 1, is characterized in that, the center of distortion point of described calculating four fault images is calculated by the mode of fitting circle, and method is as follows:
(1) often going and often arranging respectively by least square fitting circle Corner-point Grids;
(2) circle of two row matchings and the circle of two row matchings is got successively to ask center of distortion point;
(3) all center of distortions point obtained is averaged as final center of distortion point.
6. a kind of automatic calibration method based on black and white lattice corners Matching according to claim 1, is characterized in that, the method for solving of described homography matrix is:
Set up and virtually get a bird's eye view video camera angular coordinate
with the angular coordinate of orthoscopic image
homography matrix L formula:
，
；
The region of order shooting is surface level, now
, then
，
；
Wherein
be the element in L, can calculate according to the Corner-point Grids coordinate of the image after four width distortion corrections and the virtual position relationship getting a bird's eye view the coordinate of panorama sketch.
7. a kind of automatic calibration method based on black and white lattice corners Matching according to any one in claim 1 or claim 6, is characterized in that, the method for described view transformation is:
，
The point in the angular coordinate system of orthoscopic image is calculated according to above formula
the virtual level coordinates system getting a bird's eye view video camera angle point after carrying out visual transform
.
8. a kind of automatic calibration method based on black and white lattice corners Matching according to claim 1, it is characterized in that, the generation method of described lookup table is, the virtual each point getting a bird's eye view panorama sketch can calculate the coordinate of the point of changing the time after distortion correction according to homography matrix, calculate this coordinate in original fault image through the inverse transformation of distortion correction again, and relation generates lookup table according to this.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210486358.0A CN103177439B (en) | 2012-11-26 | 2012-11-26 | A kind of automatic calibration method based on black and white lattice corners Matching |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210486358.0A CN103177439B (en) | 2012-11-26 | 2012-11-26 | A kind of automatic calibration method based on black and white lattice corners Matching |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103177439A CN103177439A (en) | 2013-06-26 |
| CN103177439B true CN103177439B (en) | 2015-10-28 |
Family
ID=48637269
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210486358.0A Expired - Fee Related CN103177439B (en) | 2012-11-26 | 2012-11-26 | A kind of automatic calibration method based on black and white lattice corners Matching |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103177439B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106952311A (en) * | 2017-03-02 | 2017-07-14 | 山东省科学院自动化研究所 | Auxiliary parking system and method based on panoramic mosaic data mapping tables |
Families Citing this family (39)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103489165B (en) * | 2013-10-01 | 2016-04-06 | 中国人民解放军国防科学技术大学 | A kind of decimal towards video-splicing searches table generating method |
| CN103679729A (en) * | 2013-12-17 | 2014-03-26 | 中国人民解放军第二炮兵工程大学 | Full-automatic camera parameter calibration method based on colored calibration board |
| CN103824288A (en) * | 2014-02-17 | 2014-05-28 | 哈尔滨工业大学 | Array image registration template for lens array |
| CN103871071B (en) * | 2014-04-08 | 2018-04-24 | 北京经纬恒润科技有限公司 | Join scaling method outside a kind of camera for panoramic parking system |
| CN103955928B (en) * | 2014-04-28 | 2017-02-01 | 华为技术有限公司 | Method and device for determining camber lens distortion parameter and electronic equipment |
| CN104091322B (en) * | 2014-05-06 | 2017-06-16 | 无锡日联科技有限公司 | The detection method of laminated lithium ion battery |
| CN103996191B (en) * | 2014-05-09 | 2016-10-26 | 东北大学 | A kind of detection method of black and white chessboard table images angle point based on Least-squares minimization |
| CN105321160B (en) * | 2014-05-27 | 2019-03-22 | 穆阳 | The multi-camera calibration that 3 D stereo panorama is parked |
| CN104008548B (en) * | 2014-06-04 | 2017-04-19 | 无锡维森智能传感技术有限公司 | Feature point extraction method for vehicle-mounted around view system camera parameter calibration |
| CN105466523B (en) * | 2014-09-12 | 2019-09-17 | 航天信息股份有限公司 | The measurement method and device of heap grain height based on single camera image |
| CN105791655A (en) * | 2014-12-19 | 2016-07-20 | 宁波舜宇光电信息有限公司 | Method for computing lens distortion of photographing module |
| CN105818746A (en) * | 2015-01-05 | 2016-08-03 | 上海纵目科技有限公司 | Calibration method and system of panoramic advanced driver assistance system |
| CN104596484A (en) * | 2015-01-30 | 2015-05-06 | 黄河水利委员会黄河水利科学研究院 | Method of measuring drift ice density in ice flood season of Yellow River |
| JP6316330B2 (en) * | 2015-04-03 | 2018-04-25 | コグネックス・コーポレーション | Homography correction |
| CN105447871A (en) * | 2015-12-04 | 2016-03-30 | 北京和众视野科技有限公司 | Distortion correction algorithm of fixed focus imaging system |
| CN105809701B (en) * | 2016-03-25 | 2019-03-08 | 成都易瞳科技有限公司 | Panoramic video posture scaling method |
| CN106373091B (en) * | 2016-09-05 | 2019-05-07 | 山东省科学院自动化研究所 | Panorama gets a bird's eye view method for automatically split-jointing, system and the vehicle of image in parking |
| CN106651957B (en) * | 2016-10-19 | 2019-07-30 | 大连民族大学 | Monocular vision object space localization method based on template |
| CN106412441B (en) * | 2016-11-04 | 2019-09-27 | 珠海市魅族科技有限公司 | A kind of video stabilization control method and terminal |
| CN108108088A (en) * | 2016-11-24 | 2018-06-01 | 深圳市优朋普乐传媒发展有限公司 | A kind of method and terminal of definite pixel position |
| CN109840897A (en) * | 2017-11-28 | 2019-06-04 | 深圳市航盛电子股份有限公司 | Vehicle panoramic method for correcting image and vehicle panoramic system |
| CN108122259A (en) * | 2017-12-20 | 2018-06-05 | 厦门美图之家科技有限公司 | Binocular camera scaling method, device, electronic equipment and readable storage medium storing program for executing |
| CN108629811B (en) * | 2018-04-04 | 2021-08-20 | 广州市安晓科技有限责任公司 | Automatic calibration method and system for panoramic view of automobile |
| CN108830810A (en) * | 2018-06-07 | 2018-11-16 | 辽宁工业大学 | A kind of fisheye image distortion correction method based on rectangular projection |
| CN109242910B (en) * | 2018-08-21 | 2021-10-08 | 电子科技大学 | Monocular camera self-calibration method based on any known plane shape |
| CN109509200B (en) * | 2018-12-26 | 2023-09-29 | 深圳市繁维医疗科技有限公司 | Checkerboard corner detection method based on contour extraction and computer readable storage medium |
| CN110769223B (en) * | 2018-12-27 | 2021-09-28 | 成都极米科技股份有限公司 | Method and device for adjusting projected image and readable storage medium |
| CN110414448B (en) * | 2019-07-31 | 2021-06-29 | 北京三快在线科技有限公司 | Image processing method, image processing device, electronic equipment and storage medium |
| CN110942482A (en) * | 2019-10-14 | 2020-03-31 | 深圳市德赛微电子技术有限公司 | Lens rapid self-calibration method and electronic equipment thereof |
| CN110827346B (en) * | 2019-11-05 | 2023-11-24 | 易视腾科技股份有限公司 | Method for obtaining mapping relation between fixed plane and camera imaging thereof |
| CN110956585B (en) * | 2019-11-29 | 2020-09-15 | 深圳市英博超算科技有限公司 | Panoramic image splicing method and device and computer readable storage medium |
| CN111353952B (en) * | 2020-01-21 | 2023-05-05 | 佛山科学技术学院 | Method for eliminating black boundary after image distortion correction |
| CN111833405B (en) * | 2020-07-27 | 2023-12-08 | 北京大华旺达科技有限公司 | Calibration and identification method and device based on machine vision |
| CN111917984A (en) * | 2020-08-13 | 2020-11-10 | 上海航天测控通信研究所 | Virtual holder and control method |
| CN111739090B (en) * | 2020-08-21 | 2020-12-04 | 歌尔光学科技有限公司 | Method and device for determining position of field of view and computer readable storage medium |
| CN112541369B (en) * | 2020-12-14 | 2022-08-05 | 合肥高维数据技术有限公司 | Information code identification method and system |
| CN112802109B (en) * | 2021-02-07 | 2024-05-10 | 的卢技术有限公司 | Method for generating aerial view panorama of automobile |
| CN113570517A (en) * | 2021-07-20 | 2021-10-29 | 杨建弘 | Method for correcting two-dimensional graph of part, part measuring method and system |
| CN114693708B (en) * | 2022-04-20 | 2024-04-02 | 太原理工大学 | Anchor mesh same-row center point matching method and system |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101324749A (en) * | 2008-07-24 | 2008-12-17 | 上海交通大学 | Method for performing projection display on veins plane |
| CN101650828A (en) * | 2009-09-07 | 2010-02-17 | 东南大学 | Method for reducing random error of round object location in camera calibration |
| CN101814186A (en) * | 2010-02-04 | 2010-08-25 | 上海交通大学 | Method utilizing curve-fitting to calibrate radial distortion of camera |
| CN102663734A (en) * | 2012-03-15 | 2012-09-12 | 天津理工大学 | Fish eye lens calibration and fish eye image distortion correction method |
| CN102750697A (en) * | 2012-06-08 | 2012-10-24 | 华为技术有限公司 | Parameter calibration method and device |
-
2012
- 2012-11-26 CN CN201210486358.0A patent/CN103177439B/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101324749A (en) * | 2008-07-24 | 2008-12-17 | 上海交通大学 | Method for performing projection display on veins plane |
| CN101650828A (en) * | 2009-09-07 | 2010-02-17 | 东南大学 | Method for reducing random error of round object location in camera calibration |
| CN101814186A (en) * | 2010-02-04 | 2010-08-25 | 上海交通大学 | Method utilizing curve-fitting to calibrate radial distortion of camera |
| CN102663734A (en) * | 2012-03-15 | 2012-09-12 | 天津理工大学 | Fish eye lens calibration and fish eye image distortion correction method |
| CN102750697A (en) * | 2012-06-08 | 2012-10-24 | 华为技术有限公司 | Parameter calibration method and device |
Non-Patent Citations (2)
| Title |
|---|
| 《双目立体视觉摄像机标定及精度分析》;林俊义等;《华侨大学学报》;20110731;第22卷(第4期);全文 * |
| 《基于Harris角点检测算法的图像拼接技术研究》;贾莹;《中国优秀硕士学位论文全文数据库 信息科技辑》;20100901(第9期);第2.1.1节第3段第1－5行, 2.2.3节第1段第1－4行, 第3.4节第1－2段, 第4章4.1节, 第4.3节第4段第4－6行、第5段第1－3行 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106952311A (en) * | 2017-03-02 | 2017-07-14 | 山东省科学院自动化研究所 | Auxiliary parking system and method based on panoramic mosaic data mapping tables |
| CN106952311B (en) * | 2017-03-02 | 2020-04-07 | 山东省科学院自动化研究所 | Auxiliary parking system and method based on panoramic stitching data mapping table |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103177439A (en) | 2013-06-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103177439B (en) | A kind of automatic calibration method based on black and white lattice corners Matching | |
| CN111369439B (en) | Panoramic all-around image real-time splicing method for automatic parking space identification based on all-around | |
| US9858639B2 (en) | Imaging surface modeling for camera modeling and virtual view synthesis | |
| CN108263283B (en) | Method for calibrating and splicing panoramic all-round looking system of multi-marshalling variable-angle vehicle | |
| Hamada et al. | Surround view based parking lot detection and tracking | |
| CN104851076B (en) | Panoramic looking-around parking assisting system and camera installation method for commercial car | |
| CN101726855B (en) | Correction method of fisheye image distortion on basis of cubic projection | |
| CN111223038A (en) | Automatic splicing method and display device for vehicle-mounted all-around images | |
| CN109087251B (en) | Vehicle-mounted panoramic image display method and system | |
| KR101592740B1 (en) | Apparatus and method for correcting image distortion of wide angle camera for vehicle | |
| KR102516326B1 (en) | Camera extrinsic parameters estimation from image lines | |
| JP6522630B2 (en) | Method and apparatus for displaying the periphery of a vehicle, and driver assistant system | |
| CN103136720A (en) | Vehicle-mounted 360-degree panorama mosaic method | |
| CN109688392A (en) | AR-HUD optical projection system and mapping relations scaling method and distortion correction method | |
| CN106444837A (en) | Obstacle avoiding method and obstacle avoiding system for unmanned aerial vehicle | |
| CN103871071A (en) | Method for camera external reference calibration for panoramic parking system | |
| CN104574289A (en) | Fish-eye image distortion correction algorithm based on ellipsoidal model | |
| CN107993263A (en) | Viewing system automatic calibration method, automobile, caliberating device and storage medium | |
| CN106994936A (en) | A kind of 3D panoramic parking assist systems | |
| CN110264395A (en) | A kind of the camera lens scaling method and relevant apparatus of vehicle-mounted monocular panorama system | |
| CN107146255A (en) | Panoramic picture error calibration method and device | |
| CN107622513A (en) | A kind of piece demarcation point detection device and viewing system automatic calibration method | |
| CN113869422B (en) | Multi-camera target matching method, system, electronic device and readable storage medium | |
| CN112802109B (en) | Method for generating aerial view panorama of automobile | |
| CN111860270B (en) | Obstacle detection method and device based on fisheye camera |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20151028 Termination date: 20191126 |