CN1412537A - Method for obtaing optical projection parameter of camera - Google Patents

Method for obtaing optical projection parameter of camera Download PDF

Info

Publication number
CN1412537A
CN1412537A CN 01136401 CN01136401A CN1412537A CN 1412537 A CN1412537 A CN 1412537A CN 01136401 CN01136401 CN 01136401 CN 01136401 A CN01136401 A CN 01136401A CN 1412537 A CN1412537 A CN 1412537A
Authority
CN
China
Prior art keywords
projection
camera
image
optical axis
geometric
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
Application number
CN 01136401
Other languages
Chinese (zh)
Other versions
CN1203297C (en
Inventor
詹国祯
张创然
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Oupuluo Science & Technology Co Ltd
Original Assignee
Oupuluo Science & Technology Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Oupuluo Science & Technology Co Ltd filed Critical Oupuluo Science & Technology Co Ltd
Priority to CN 01136401 priority Critical patent/CN1203297C/en
Publication of CN1412537A publication Critical patent/CN1412537A/en
Application granted granted Critical
Publication of CN1203297C publication Critical patent/CN1203297C/en
Anticipated expiration legal-status Critical
Expired - Fee Related legal-status Critical Current

Links

Images

Landscapes

  • Image Processing (AREA)

Abstract

The method for obtaining optical projection parameter of camera includes the following steps: utilizing the characteristics of scrambed image which has centrosymmetry, using a correcting article with several symmetrical and concentric geometrical figures to define the optical axis of camera lens, and further find out its projection centre point on the optical axis by using error testing method, at the same time solving its equivalent focusing and projection mode. Said invention is simple, can accurately define internal and external parameters of camera, so that it has extensive application range.

Description

Method for calculating optical projection parameters of camera
Technical Field
The present invention relates to a method for processing and displaying digital images, and more particularly, to a method for restoring a distorted image to a normal perspective image by finding optical projection parameters such as a projection center (viewpoint) and an equivalent focal length (focal length) of a camera lens.
Background
It is often desired to use an optical lens to image, and it is desired that the imaging result of linear projection is satisfied, and the image captured by the user under the similar lens is very close to the real world, but it has a disadvantage of small visual angle, for example, the visual angle is about 45-55 degrees in the case of a standard lens, so that the wide-angle lens or the fish glasses are adopted in the field such as a monitor or an endoscope requiring a large visual angle, so as to capture the image with a large visual angle at one time.
The fisheye lens is used for capturing very wide-angle images, and a camera equipped with the fisheye lens can capture 180 degrees or more of visual field and capture an image once without moving the camera itself for multiple times of shooting, but the original image has serious distortion problem with the increase of the visual angle, so that the image correction technology is usually needed to output perspective images closer to the real world. The accuracy of image correction affects the range to which the fisheye lens can be applied; in general monitoring systems, if only the movement direction of people or objects in a monitoring range needs to be seen, the situation of picture distortion can be tolerated; if the image is a picture for shooting Virtual Reality (Virtual Reality), the image only needs to be normally and clearly seen; however, if the method is applied to the field of image positioning, such as endoscope or robot vision, it is difficult to accurately determine the real position of the image in space under the condition that many optical parameters of the current fish-eye lens are unknown.
In any case, for consumers, it is most competitive in the market if the lens has a large viewing angle, a correct and clear image and an accurate positioning. Moreover, the fisheye lens has the advantage of infinite depth of field, which is incomparable with other types of lenses, so that how to correct the image generated by the fisheye lens is an important subject.
In the prior art, many methods for correcting the distorted image have appeared. R.y.tsai (1987) uses five points known to be not coplanar in space and the operation of a rotation matrix and a transformation matrix to calculate parameters such as the projection center (viewpoint) and the focal length (focal length) of the lens, and although the result calculated by the simulation is quite accurate, the theoretical derivation is based on linear projection, and when applied to a severe nonlinear projection mechanism such as a fisheye lens, the error of calculating the equivalent focal length is quite large, and therefore, the method cannot be directly used as a general correction method. Then, a simple calibration method for the fisheye lens appears, please refer to fig. 1A and 1B, wherein fig. 1B corresponds to the hemispherical space projection mode of fig. 1A, assuming that the image plane 1 is an ellipse (or a circle), the intersection 13 of the major axis 11 and the minor axis 12 is exactly the image center (point C in the figure), and it is also believed that the field of view captured by the fisheye lens is exactly 180 degrees, based on the above-mentioned several premises, it is inferred that the incident azimuth (zenithal angle) at the edge of the image plane 1 is pi/2, the azimuth of the center point 13 is 0, and the remaining points individually infer the azimuth θ according to their relative positions with the point C13 and the edge. For example, point a in fig. 1A should correspond to point a in fig. 1B. The above-mentioned correction method for fisheye lens image is simple and does not need additional correction object (calibration target) for assistance. This is a camera description model used by many fisheye cameras that directly correct for flat images. However, the assumption premise of the method is quite uncertain, firstly, the distortion central point 13 of the picture is not necessarily exactly the actual central point of the image, and the edge of the fisheye lens image is fuzzy, so that the boundary of 180 degrees is difficult to determine; most importantly, the radius and center of the hemisphere cannot be located, and the fidelity of the corrected image is difficult to evaluate in this image-based analysis mode. Obviously, such a correction method is not suitable for the field of stereoscopic image localization.
In the disclosure of the patent, TeleRobotic International inc. proposes several patents (U.S. patent nos. 5185667, 5313306, 5359363, 5384588) related to fisheye lens cameras, and its technical content is to use a fisheye lens to capture the whole hemispherical field of view, then use a computer to control a specially designed electronic circuit, and convert the original distorted image into a normal perspective image in a digital conversion manner and display it on a screen according to the aforementioned camera mode. Subsequently, Interactive Pictures Corporation proposed a series of new and improved solutions in accordance with the aforementioned technology of telerobotic international inc (U.S. patent nos. 5764276, 5877801, 5903319, 5990941, 6002430, 6147709, 6201574B 1). However, in any of the above technical solutions, the image is corrected by using a fixed projection mode, and the parameters of the fisheye lens camera, namely, the projection center (viewpoint) and the focal length (focal length), are not exactly determined, and the projection mode, namely, the "equal distance projection" (equi-distance projection), which is the most common fisheye lens projection mode, is assumed as described in fig. 1A and 1B. Under the conditions that the precise parameters are not determined and the projection mode is unknown, only the original image can be expanded and restored, the accuracy of the corrected image cannot be considered, and the application in the field of positioning the stereoscopic image is necessarily limited.
In fact, in the presently known fisheye lens Projection modes, in addition to the "Equidistant Projection" (equi-distance Projection), there are also "stereoscopic Projection" (stereo Projection) and "orthogonal Projection" (orthogonal Projection), and the mathematical relations of the optical projections are respectively as follows:
1. equidistant projection: IH ═ f θ
2. Projection of stereoscopic images: IH 2f × tan (θ/2)
3. Right-angle projection: IH ═ f × sin θ
Wherein,
IH: the distance (image height) between the image point and the optical axis of the lens;
f: equivalent focal length of the fisheye lens;
θ: the incident angle of the object point to the focal plane is the angle between the incident light and the optical axis.
Theoretically, "stereoscopic image projection" is the best fisheye lens projection mode; however, for optical design reasons, the fisheye lens conforming to the theory of "equal distance projection" is the easiest to manufacture, which also makes the currently known correction techniques have a hypothesis: the projection mode of all fisheye lenses is "equal distance projection", but in fact it is not necessary.
On the other hand, although the lens is designed according to a certain design during manufacturing, it is difficult to verify whether the original design specification is maintained after manufacturing; in addition, the effective focal length and the observable viewing angle of the fisheye lens may vary after the fisheye lens is assembled into a real system (e.g., a camera), so that if a simple and versatile technique is available, the specifications of the fisheye lens or the image capture device comprising the fisheye lens can be checked, so that the fisheye lens has more definite specifications when shipped, and the added value of the fisheye lens can be greatly increased.
For fisheye lenses, one of ordinary skill in the art would consider there to be no "real" center of projection, which is in the perspective of a straight line projection. If the corresponding projection mode can be concluded and the optical projection parameters such as the projection center and the equivalent focal length can be found, the correct projection image can be easily corrected, and the method can be applied to the field of three-dimensional image positioning and product quality management.
Disclosure of Invention
In view of the above, the present invention provides a solution to the image correction problem of the severe non-linear projection mechanism based on the identification of the lens-based native optical projection mode.
Another objective of the present invention is to determine the projection center and effective focal length of the camera, so that the camera can more accurately correct the image and can be applied in the field of stereo image determination. It is still another object of the present invention to provide a method for inspecting the spatial projection mechanism of a fisheye lens or a photographing device thereof for quality control before product specification or shipment.
In accordance with the above-described objects of the present invention, a method for determining optical projection parameters of a camera is provided. Firstly, a plane-form correcting object with a test pattern is placed in the visual field of a fisheye lens camera, the direction and the position of the correcting object are adjusted, so that the corresponding image formed by the correcting object is also a similar geometric pattern, the center of the test pattern and the center of the corresponding image are connected at the moment, and the optical axis of the fisheye lens can be determined; then, a point is sought along the optical axis by trial and error method, so that the corresponding image height of the test pattern and the angle of orientation (null angle) between the sought point and the test pattern are in accordance with a projection mode.
The Projection mode may be one of currently known fisheye lens Projection modes such as "equal distance Projection" (equi distance Projection), "stereo Projection" (stereo Projection), or "orthogonal Projection" (orthogonal Projection). The found fixed point is the projection center (viewpoint) of the fisheye lens camera, and the equivalent focal length can be obtained from the mathematical relation of the projection mode.
The invention can accurately determine the projection center and the equivalent focal length of the fisheye lens camera and can find the original projection mode, thereby easily correcting the distorted image and finding out the projection curve of the visual field space.
In order to make the aforementioned and other objects, features and advantages of the present invention comprehensible, preferred embodiments accompanied with figures are described in detail below.
Drawings
FIGS. 1A and 1B are schematic diagrams illustrating a conventional fisheye lens image correction method based on a planar image and a transformation of the spatial projection thereof;
FIG. 2 is a schematic diagram of an optical path defining an optical axis of a fisheye lens according to the present invention;
FIGS. 3A-3D illustrate four embodiments of the invention for calibrating test patterns on an object;
FIG. 4 is a schematic diagram of an optical path for finding out a projection center and an equivalent focal length in the "equal distance projection" mode according to the present invention;
FIG. 5 is a graph showing an asymptotic curve of the center of projection obtained in the actual test of the present invention; and
FIG. 6 shows an embodiment of the present invention for solving the problem of the calibration object with a viewing angle greater than 180 degrees.
Detailed Description
Although a lens having axisymmetric nonlinear distortion, such as a fisheye lens, projects an image with serious distortion problems, the distortion phenomenon has the following characteristics: the present invention firstly uses the phenomenon to define the optical axis of the fisheye lens, then uses the optical axis as reference to find the projection center of the fisheye lens camera, and can calculate the focal length and summarize the projection mode.
Referring to fig. 2, to implement the method of the present invention, a calibration object (calibration target)22 is utilized, and the calibration object 22 has at least one geometric figure, and if there are a plurality of geometric figures, the geometric figures are concentric and symmetrical, such as concentric circles shown in the figure, which are called test patterns 220. During the calibration, the calibration object 22 is placed in the field of view of a fisheye lens camera, and the test pattern 220 forms a corresponding image 230 at an image plane 23 behind the fisheye lens 24. According to the feature that the distorted image is centrosymmetric, if the test pattern 220 and the image plane 23 are parallel and the center point is aligned with the optical axis 21 of the fisheye lens 24, the corresponding image 230 is also composed of concentric circles as with the test pattern 220, so that the centers of the circles connecting the test pattern 220 and the corresponding image 230, that is, the optical axis 21 of the fisheye lens 24, are connected as long as the position and the direction of the calibration object 22 are properly adjusted until the formed corresponding image 230 is composed of concentric tracks.
The test patterns 220 that can be used in the method of the present invention are not limited to the concentric circles shown in fig. 3A, but the test patterns 220 are preferably composed of concentric and symmetrical geometric figures, and besides the concentric circles, the test patterns 220 that can be used in the method of the present invention can also be concentric squares as shown in fig. 3B, concentric triangles as shown in fig. 3C, or concentric hexagons as shown in fig. 3D; even combining any number of concentric and symmetrical circles, squares, triangles and polygons is another possible embodiment. However, in the actual test, in consideration of the problem of the tubular distortion of the image 230, in addition to the circular shape, it is necessary to set the vertex of the geometric figure (such as triangle or square) as a feature coordinate (functional coordinates) as a reference point for the calibration process.
Referring to fig. 4, after the optical axis 21 of the fisheye lens 24 is determined, it can be known from the optical theory that: the center of projection of the lens 24 is always located at a certain point on the optical axis 21, and is often located within the fisheye lens 24, so that the search range is greatly reduced, and therefore, the present invention uses a trial and error method to test one by one along the optical axis 21 within the fisheye lens 24 to find the center point of projection of the fisheye lens 24. As for the testing method, also using the test pattern 220 on the calibration object 22, if the "equal distance Projection" (equal distance Projection) mode is taken as an example, assuming that a fixed point 241 on the optical axis 21 is determined as the optical origin, the direction angle θ i of the connecting line of the fixed point 241 to the object point 221 on each concentric circle deviating from the optical axis 21 and the distance IH of the image point 231 corresponding to the object point 221 to the optical axis 21 (in this case, the image center) can be measuredi(or called image height), with this data, we can get the theta from each concentric circleiAnd IHiTo obtain fi. If the camera is completely in accordance with the relation of 'equidistant projection', f calculated by concentric circles with different radii is distortediIs a constant.
In practice, the test pattern 280 is set to twenty concentric circles, and each adjacent concentric circle is separated by 5 mm. For convenience of description, it is assumed that the intersection of the center point of the calibration object 22 and the optical axis 21 is (0, 0, 0) and the optical axis is the z-axis, which can be expressed as (0, 0, z), if the distance between the projection center and the calibration object 22 is D, the distance between the projection center and each concentric circle on the calibration object 22 is setRadius riCorresponding to each image height IHiDue to IHiAnd thetaiAre all functions of D, so that the "equal distance projection" can be changed to the following pattern IHi(D)=fθi(D) Where i is 1-20, the outermost circular pattern may be taken as a reference, IH20(D)=fθ20(D) After simple operation, the following equation can be obtained:
IHi(D)/IH20(D)-θi(D)/θ20(D)=0
if (O, O, D) is changed to an arbitrary point on the z-axis, an error relation is obtained as follows:
ei(z)=IHi(z)/IH20(z)-θi(z)/θ20(z)
according to the above formula, the test point is at (O, O, D), ei(z) is a minimum, the distance of the calibration object 22 can be fixed.
However, the above formula only takes two concentric circles, and the calculated result is to consider the effective field of view of the testing range covering camera and observe the distortion of the concentric circle imaging along the radial direction, so a weight function w is set with reference to the covering range of each circular tracki(D)=(IHi(D)-IHi-i(D))/IH20(D) Wherein IH0(D) 0, to treat the success of each trajectory fairly. Therefore, in the fitting process of the present invention for finding the projection center on the optical axis 21, the applied error function is: <math> <mrow> <mi>&epsiv;</mi> <mrow> <mo>(</mo> <mi>z</mi> <mo>)</mo> </mrow> <mo>=</mo> <munderover> <mi>&Sigma;</mi> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mn>20</mn> </munderover> <mi>abs</mi> <mrow> <mo>(</mo> <msub> <mi>e</mi> <mi>i</mi> </msub> <mrow> <mo>(</mo> <mi>z</mi> <mo>)</mo> </mrow> <mo>&times;</mo> <msub> <mi>w</mi> <mi>i</mi> </msub> <mrow> <mo>(</mo> <mi>D</mi> <mo>)</mo> </mrow> <mo>)</mo> </mrow> </mrow> </math>
where z is an arbitrary point on the optical axis 21, if a unique point can be found so that epsilon (z) is the minimum or approaches to 0, then the point is the projection center of the fisheye lens camera. As for focal length f, according to the measured IHi(D) And relative theta thereofi(D) On the basis, it is calculated using the following formula: <math> <mrow> <mi>f</mi> <mrow> <mo>(</mo> <mi>D</mi> <mo>)</mo> </mrow> <mo>=</mo> <munderover> <mi>&Sigma;</mi> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mn>20</mn> </munderover> <msub> <mi>f</mi> <mi>i</mi> </msub> <mrow> <mo>(</mo> <mi>D</mi> <mo>)</mo> </mrow> <mo>&times;</mo> <msub> <mi>w</mi> <mi>i</mi> </msub> <mrow> <mo>(</mo> <mi>D</mi> <mo>)</mo> </mrow> </mrow> </math>
wherein f isi(D)=IHi(D)/θi(D) If the lens completely conforms to the projection mode, the measurement is error-free, and the value D is found to be accurate, then f (D) should be equal to any of fi(D) This is the focal length f of the lens. In fact, each f obtainedi(D) The statistical standard deviation of (2) can be further utilized to estimate the accuracy of the projection mode, that is, the following formula can be used as an index of the degree of adaptation to the "equidistant projection" mode: <math> <mrow> <mi>&sigma;</mi> <mrow> <mo>(</mo> <mi>D</mi> <mo>)</mo> </mrow> <mo>=</mo> <mrow> <mo>(</mo> <munderover> <mi>&Sigma;</mi> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mn>20</mn> </munderover> <msup> <mrow> <mo>(</mo> <msub> <mi>f</mi> <mi>i</mi> </msub> <mrow> <mo>(</mo> <mi>D</mi> <mo>)</mo> </mrow> <mo>-</mo> <mi>f</mi> <mrow> <mo>(</mo> <mi>D</mi> <mo>)</mo> </mrow> <mo>)</mo> </mrow> <mn>2</mn> </msup> <mo>)</mo> </mrow> <mo>/</mo> <mrow> <mo>(</mo> <mn>20</mn> <mo>-</mo> <mn>1</mn> <mo>)</mo> </mrow> </mrow> </math>
referring to fig. 5, there is shown an asymptotic curve for D along the z-axis when tested with a BV-7112 camera manufactured by opulmo, which is mounted with a DW9813 fisheye lens manufactured by Daiwon Optical (Korea), and which has a focal length of 1.78mm and a diagonal viewing angle of 170 degrees according to the manufacturer's specification. The first test result is shown in the figure by a solid line, and the calibration object 22 is moved outward by 5mm (dD 5), 10mm (dD 10), 15mm (dD 15), 20mm (dD 20) and 25mm (dD 25) respectively based on the D value under the test (dD 0), and the same test is performed, and it is found that in all six cases, there is a very clear minimum value of epsilon (z), which proves that the center of projection of the fisheye lens is definite and can be found out very accurately by the present invention. The test data associated with fig. 5 is shown in table 1:
TABLE 1 parameters and results of the tests
dD
0 5 10 15 20 25
D 14.7 19.6 25.2 30 35.3 39.5
f(D) 1.805 1.788 1.827 1.796 1.798 1.784
σ(D) 0.005 0.002 0.0091 0.0058 0.0062 0.0052
(unit: mm)
Looking at the data in Table 1, regardless of which of the six positions the calibration object 22 is at, the calculated focus landing points (f (D) ± σ (D)) are very close and σ (D) is relatively small, indicating that the method of the present invention has relatively good accuracy and precision.
When the projection center and the focal length of the fisheye lens camera are derived, the image correction becomes very easy. In the case of a fisheye lens camera conforming to the "equal distance projection" mode, please refer to fig. 4 again, in which the mapping relationship between the object point 221 of the hemispherical field of view and the corresponding image point 231 is very direct, and the image plane can be represented by a spherical surface 25 with the focal length as the radius length.Thus, logically, the incident ray focal plane can have the same incident angle and refraction angle as the linear projection. The length of the projected image of any line of sight can be directly expressed by the arc length of its incident angle (i.e., the reflection angle). So that the image length is IH ═ f θ at the incident angle θ (i.e., the reflection angle)iThe image height of the corrected image point 231 'is IH' ═ f × tan θ. As is clear from FIG. 4, Δ OPQ is similar to Δ OP ' Q ', so that each IH ' that is corrected corresponds to riIn equal proportion, the method can restore the distorted image very accurately.
In the testing process, the method of the present invention simultaneously checks the projection mode of the obtained data to minimize the error value epsilon (z), so that the method of finding the projection center and the focal length in the present invention is not limited to be applied to a fisheye lens camera conforming to the "equidistant projection" mode, but can also be applied to a camera conforming to the "stereoscopic image projection" (IH ═ 2f × tan (θ/2)) or "orthogonal projection" (IH ═ f × sin θ) mode, and even can identify any nonlinear lens with known projection function, and has the function of summarizing and determining the real projection mode of each camera. The types of cameras to which the present invention can be applied include CCD cameras, CMOS cameras, digital cameras, or a conventional camera using a film.
The planar calibration object 22 cannot reach a view angle of 180 degrees, and the radius of the calibration object 22 is infinite when the view angle is 180 degrees, but the fisheye lens can capture the view angle. As to how to solve this problem, the present invention also provides a solution, as shown in fig. 6, by extending the calibration object 22 into a hollow cylinder, such as an open-ended iron can, the bottom of which is drawn with the calibration object 22 in a planar form as described above, and is used to calibrate the camera position and distance. When it has been located by the camera distance D, the optical axis 21 and the central axis of the ring coincide, and the lens extends from the surrounding surface 22a of the cylinder by the distance D to the camera horizontal line of the fish-eye lens, as shown in the figure, an object point 222 located on the horizontal plane defined by the projection center 241 is shown, and its corresponding image point 232 is located on the 180-degree boundary of the image. The extension mode can be applied to the lens with the visual field angle larger than 180 degrees.
Since the invention can exactly determine the projection center and equivalent focal length of the fisheye lens camera, if the orientation of the calibration object 22 is referred (the orientation of the calibration object 22 is available), a plurality of sets of cameras can be relatively positioned to form a stereo vision system, and the system can have a larger operable visual angle than a common lens. Locating the three-dimensional orientation of an object by triangulation is well known to those skilled in the art, and thus, a detailed description thereof will not be provided herein. However, the focal plane incident angle can be quickly obtained from the 'so-called' distorted image height through the projection formula (because the calculation of the positioned point, the incident angle of the stereo vision system camera set is the necessary step for stereo vision three-dimensional positioning), which reduces many calculation steps compared with the method that the image height corresponding to linear projection is corrected by a non-linear high-order function and then the incident angle is calculated by combining the arctan function of the linear equivalent focal length. Therefore, the description of a severe nonlinear projection mechanism like a fish-eye lens by using a native projection mode has an absolute advantage.
The method for positioning the fisheye lens camera and solving the equivalent focal length of the fisheye lens camera provided by the invention has the following advantages:
1. the projection center of the fisheye lens camera can be determined accurately and the focal length of the fisheye lens camera can be obtained, so that the distorted image can be easily restored to a normal image under a central projection mechanism.
2. The real projection pattern of the fisheye lens camera can be generalized or found.
3. The method can be applied to the field of three-dimensional image positioning, and has a simpler and faster incident angle operation mode.
4. The correction method is simple and low in cost, and is suitable for any projection mode of fisheye lens or camera with non-linear projection mechanism.
Although the present invention has been described with reference to a preferred embodiment, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.

Claims (21)

1. A method for determining optical projection parameters of a camera for obtaining a projection center and an equivalent focal length of the camera, wherein the applied camera has a lens with an axisymmetric nonlinear distortion, the method comprising:
providing a calibration object having a test pattern, wherein the test pattern is composed of at least one geometric figure;
aligning the center of the test pattern to an optical axis of the lens; and
seeking a fixed point along the optical axis so that at least one image height corresponding to the geometric figure formed by the geometric figure on an image plane and an included angle between the incident light of the geometric figure and the optical axis accord with a projection mode.
2. The method of claim 1, wherein the geometric figure is one selected from the group consisting of a circle, a square, a triangle, and a polygon.
3. The method of claim 1, wherein the geometric figure is a plurality of concentric circles, concentric squares, concentric triangles or concentric polygons.
4. The method of claim 1, wherein the step of centering the test pattern on the optical axis of the camera further comprises:
placing the calibration object in front of the camera to form an image of the test pattern on the image plane; and
the position of the calibration object is moved so that the image is similar to the test pattern, and the center connecting the test pattern and the image is the optical axis.
5. The method of claim 1, wherein the axisymmetric nonlinear anamorphic lens is a fisheye lens.
6. The method of claim 1 wherein the projection mode is a known non-linear projection mode.
7. The method of claim 1, wherein the projection mode is one selected from a group consisting of an equal distance projection, a right angle projection and a stereoscopic projection.
8. The method of claim 1, wherein the fixed point is a projection center of the camera lens, and the image height of the at least one geometric figure and an angle between the incident light and the optical axis are calculated by substituting the image height and the angle into the projection mode to obtain an equivalent focal length of the camera lens.
9. The method of claim 1, wherein the calibration object has at least one flat surface for providing the test pattern.
10. The method as claimed in claim 9, wherein a surrounding surface is further extended vertically from the periphery of the plane, so that the calibration object is a hollow pot with one open end.
11. The method of claim 1, wherein the camera is one selected from the group consisting of a CCD camera, a CMOS camera, a digital camera, and a conventional film camera.
12. A method for obtaining optical projection parameters of a camera, which is used for obtaining an optical axis, a projection center and an equivalent focal length of the camera, wherein the applied camera has a lens with axisymmetric nonlinear distortion, the method comprises:
providing a calibration object having a test pattern, wherein the test pattern is composed of at least one geometric figure;
placing the calibration object in the field of view in front of the lens to form a corresponding image on an image plane by the test pattern;
moving the position of the calibration object to make the corresponding image similar to the test pattern, and connecting the center of the test pattern and the center of the image as an optical axis; and
a fixed point is sought along the optical axis such that at least one image height corresponding to the geometric figure and an angle between incident light of the geometric figure and the optical axis conform to a projection pattern.
13. A method for determining optical projection parameters for a camera as claimed in claim 12, wherein said geometric figure is one selected from the group consisting of a circle, a square, a triangle and a polygon.
14. The method of claim 12, wherein the geometric figure is a plurality of concentric circles, concentric squares, concentric triangles or concentric polygons.
15. A method as claimed in claim 12, wherein the axisymmetric non-linearly anamorphic lens is a fisheye lens.
16. The method of claim 12, wherein the projection pattern is a known non-linear projection pattern.
17. The method of claim 12, wherein the projection mode is one selected from the group consisting of an equal distance projection, a right angle projection and a stereoscopic projection.
18. The method of claim 12, wherein the fixed point is a projection center of the camera lens, and the image height of the at least one geometric figure and an angle between the incident light and the optical axis are calculated by substituting the image height of the at least one geometric figure and the angle between the incident light and the optical axis into the projection mode, so as to obtain an equivalent focal length of the camera lens.
19. The method as claimed in claim 11, wherein the calibration object has at least one flat surface for providing the test pattern.
20. The method of claim 19, wherein a surrounding surface is further extended vertically from the periphery of the plane, so that the calibration object is a hollow pot with one open end.
21. The method of claim 12, wherein the camera is one selected from the group consisting of a CCD camera, a CMOS camera, a digital camera, and a conventional film camera.
CN 01136401 2001-10-15 2001-10-15 Method for obtaing optical projection parameter of camera Expired - Fee Related CN1203297C (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN 01136401 CN1203297C (en) 2001-10-15 2001-10-15 Method for obtaing optical projection parameter of camera

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN 01136401 CN1203297C (en) 2001-10-15 2001-10-15 Method for obtaing optical projection parameter of camera

Publications (2)

Publication Number Publication Date
CN1412537A true CN1412537A (en) 2003-04-23
CN1203297C CN1203297C (en) 2005-05-25

Family

ID=4673641

Family Applications (1)

Application Number Title Priority Date Filing Date
CN 01136401 Expired - Fee Related CN1203297C (en) 2001-10-15 2001-10-15 Method for obtaing optical projection parameter of camera

Country Status (1)

Country Link
CN (1) CN1203297C (en)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101319896B (en) * 2007-06-08 2010-10-13 中国科学院沈阳应用生态研究所 Method for measuring three-dimensional structure of forest gap
CN101726855B (en) * 2009-11-13 2011-05-11 河北工业大学 Correction method of fisheye image distortion on basis of cubic projection
CN102256053A (en) * 2010-05-18 2011-11-23 鸿富锦精密工业(深圳)有限公司 Image correcting system and method
CN102316242A (en) * 2010-07-06 2012-01-11 鸿富锦精密工业(深圳)有限公司 Image correcting system and correcting method thereof
CN108734666A (en) * 2017-04-13 2018-11-02 杭州海康威视数字技术股份有限公司 A kind of fisheye image correcting method and device

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101319896B (en) * 2007-06-08 2010-10-13 中国科学院沈阳应用生态研究所 Method for measuring three-dimensional structure of forest gap
CN101726855B (en) * 2009-11-13 2011-05-11 河北工业大学 Correction method of fisheye image distortion on basis of cubic projection
CN102256053A (en) * 2010-05-18 2011-11-23 鸿富锦精密工业(深圳)有限公司 Image correcting system and method
CN102316242A (en) * 2010-07-06 2012-01-11 鸿富锦精密工业(深圳)有限公司 Image correcting system and correcting method thereof
CN108734666A (en) * 2017-04-13 2018-11-02 杭州海康威视数字技术股份有限公司 A kind of fisheye image correcting method and device
CN108734666B (en) * 2017-04-13 2021-03-26 杭州海康威视数字技术股份有限公司 Fisheye image correction method and device

Also Published As

Publication number Publication date
CN1203297C (en) 2005-05-25

Similar Documents

Publication Publication Date Title
Zhang et al. A robust and rapid camera calibration method by one captured image
JP6858211B2 (en) Devices and methods for positioning a multi-aperture optical system with multiple optical channels relative to an image sensor.
JP6967715B2 (en) Camera calibration method, camera calibration program and camera calibration device
US6985183B2 (en) Method for exploring viewpoint and focal length of camera
CN109859272B (en) Automatic focusing binocular camera calibration method and device
CN106056620B (en) Line laser camera measurement system calibrating method
US8619248B2 (en) System and method for calibrating ultra wide-angle lenses
CN112562014B (en) Camera calibration method, system, medium and device
CN115830103A (en) Monocular color-based transparent object positioning method and device and storage medium
CN109961485A (en) A method of target positioning is carried out based on monocular vision
CN107507244A (en) Camera calibration method, proving operation method and the caliberating device of a kind of single-frame images
US10628968B1 (en) Systems and methods of calibrating a depth-IR image offset
CN113119129A (en) Monocular distance measurement positioning method based on standard ball
CN108550171B (en) Linear array camera calibration method containing eight-diagram coding information based on cross ratio invariance
CN113920206A (en) Calibration method of perspective tilt-shift camera
CN114998448A (en) Method for calibrating multi-constraint binocular fisheye camera and positioning space point
CN112950727B (en) Large-view-field multi-target simultaneous ranging method based on bionic curved compound eye
CN114926538A (en) External parameter calibration method and device for monocular laser speckle projection system
JP3696336B2 (en) How to calibrate the camera
TW565736B (en) Method for determining the optical parameters of a camera
JP3704494B2 (en) How to check camera viewpoint and focal length
CN1412537A (en) Method for obtaing optical projection parameter of camera
CN110163922B (en) Fisheye camera calibration system, fisheye camera calibration method, fisheye camera calibration device, electronic equipment and storage medium
CN109682312B (en) Method and device for measuring length based on camera
CN111754584A (en) Remote large-field-of-view camera parameter calibration system and method

Legal Events

Date Code Title Description
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
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
C19 Lapse of patent right due to non-payment of the annual fee
CF01 Termination of patent right due to non-payment of annual fee