WO2013081322A1 - Procédé d'estimation d'informations de vol d'objet sphérique à l'aide de marqueur en cercle - Google Patents

Procédé d'estimation d'informations de vol d'objet sphérique à l'aide de marqueur en cercle Download PDF

Info

Publication number
WO2013081322A1
WO2013081322A1 PCT/KR2012/009747 KR2012009747W WO2013081322A1 WO 2013081322 A1 WO2013081322 A1 WO 2013081322A1 KR 2012009747 W KR2012009747 W KR 2012009747W WO 2013081322 A1 WO2013081322 A1 WO 2013081322A1
Authority
WO
WIPO (PCT)
Prior art keywords
circle
circle marker
image
spherical object
marker
Prior art date
Application number
PCT/KR2012/009747
Other languages
English (en)
Korean (ko)
Inventor
정미애
Original Assignee
Jung Mi-Ae
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 Jung Mi-Ae filed Critical Jung Mi-Ae
Publication of WO2013081322A1 publication Critical patent/WO2013081322A1/fr

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/20—Analysis of motion
    • G06T7/246—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments
    • G06T7/248—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments involving reference images or patches
    • A—HUMAN NECESSITIES
    • A63—SPORTS; GAMES; AMUSEMENTS
    • A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
    • A63B69/00—Training appliances or apparatus for special sports
    • A63B69/36—Training appliances or apparatus for special sports for golf
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/70—Determining position or orientation of objects or cameras
    • G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
    • G06T7/74—Determining position or orientation of objects or cameras using feature-based methods involving reference images or patches
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30204—Marker
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30221—Sports video; Sports image
    • G06T2207/30224—Ball; Puck
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30241—Trajectory

Definitions

  • the present invention relates to a method for estimating the flight information of the spherical object, that is, the flight speed (speed and direction of flight) and the rotational speed (speed and direction of rotation axis) from a mono image or stereo image of a flying spherical object acquired at two or more time points. It is about.
  • Korean Patent No. 10-0784967 is for estimating golf club hit information and golf ball flight information, using four multiple exposure cameras and a laser-based trigger device of a transmission and reception method, and using a multiple exposure of a multiple exposure camera. Images of golf balls are taken at various points in time. In order to estimate the speed of the golf club, three reflection points are attached to the golf club. The rotation information of the golf ball is determined by the angle between the tangential straight lines of the striped arc at two points at the intersection of two striped arcs drawn on the golf ball. It was calculated and estimated by the rotation angle. However, this rotation angle estimation method is not only an approximate method of inferior accuracy, but also has a problem in that the golf ball must be placed so that the intersection point is always visible.
  • Korean Patent No. 10-0871595 relates to a method for estimating spherical object flight information using one line scan camera and two high speed cameras.
  • the initial velocity of the spherical object from the line image of the spherical object passing through the line scan camera The stereo images were obtained by taking a picture so that spherical objects at three time points do not overlap one image with a high-speed camera capable of multiple exposures by controlling the multiple exposure time intervals based on the predicted initial speed.
  • the center point of the spherical object was calculated from the stereo images captured by this method, and the rotation information of the spherical object was estimated by calculating three intersection points of the line markers engraved on the spherical object from the multi-exposure image.
  • the rotation information estimation method using three intersection points of the line markers used in the patent is much lower in accuracy by calculating the rotation speed and the rotation axis with only three intersection points, and if one of the three is not found, the calculation becomes impossible.
  • Korean Patent No. 10-0937922 two pairs of stereo images are acquired by using four trigger cameras and a general camera for acquiring high speed cameras to avoid the use of one trigger camera and a high speed camera informing the point of image acquisition.
  • the rotation speed and rotation axis of the golf ball were estimated.
  • the patent uses a number of marking points engraved on the golf ball to estimate the rotational information of the golf ball, but not only the golf ball should be placed on the hitting mat so that the marking points are easily visible to the camera, the golf ball is a strong hitting As it is repeated, the marking points of the golf ball are more likely to be erased than the line markers, and thus the reliability is low.
  • Korean Patent No. 10-1019823 acquires a multiple exposure image of a moving object using two cameras and a strobe device, extracts the image of the object from the acquired image, and obtains coordinates of the center points of the object images.
  • the motion trajectory of the object was obtained using the coordinates, but the rotation information of the flying object could not be estimated.
  • Korean Patent No. 10-1044887 obtained an image of a golf ball by using one line scan camera and two high speed cameras as a trigger, and estimated the flight speed of the golf ball by using the acquired image, but the rotation speed was Can not be estimated.
  • the intersection point of the line marker, the line marker angle at the intersection point, or the positions of several marking points are used.
  • the estimation method is not only accurate, but also robust to the effects of lighting, image noise, and the like.
  • Patent Document 1 KR 10-0784967 B1 2007. 12. 11.
  • Patent Document 2 KR 10-0871595 B1 2008. 11. 28.
  • Patent Document 3 KR 10-0937922 B1 January 13, 2010
  • Patent Document 4 KR 10-1019823 B1 2011. 2. 25.
  • the method of estimating the rotational speed of a spherical object using three intersections of line markers has a much lower accuracy by calculating the rotational speed with only three intersections, and it is impossible to calculate when one of the three is not found.
  • the method of estimating the rotational speed of a spherical object by using several marking points requires not only to place the golf ball on the hitting mat so that the marking points are clearly visible to the camera, but also to the golf ball and the baseball ball. As it is repeated, the marking points are more likely to be erased, which results in a lower reliability.
  • the flight information estimation result of the spherical object by using the geometric characteristics of the circle markers without using the positional information in the image of the points such as the intersection of the line marker or the marking point to increase the accuracy and reliability of the.
  • the spherical object flight information estimation method using the present circle marker extracts the original marker image point from the camera image at different viewpoints, and at different viewpoints derived from the extracted original marker image point.
  • the position vector and normal vector of the circle marker are obtained from the elliptic equation of, and the flight speed and the rotation speed of the spherical object are calculated.
  • the present invention uses the geometrical characteristics of the circle marker instead of using the location information in the image of the point such as the intersection of the line marker or the marking point to estimate the flight information, thereby increasing the accuracy of the flight information estimation result, and using a plurality of circle markers.
  • the average of the results calculated from the circle markers can be used, thereby increasing the robustness to the effects of image noise and lighting.
  • the flight information of spherical objects can be estimated even when the circle markers engraved on the spheres are partially erased for strong hitting.
  • the rotation speed can be calculated using only the mono image of the flying sphere. If the size of the sphere is not known in advance, the flight of the sphere is performed from the stereo image of the flying sphere. Information can be estimated.
  • FIG. 1 is a conceptual diagram showing a relationship between a circle image in a three-dimensional space and a circle taken on a single camera image plane.
  • FIG. 2 is an exemplary view illustrating that various shapes of a circle may appear on a single camera image surface according to a relative posture between a circle and a single camera in FIG. 1.
  • FIG 3 is an exemplary view of a circle marker marked on a spherical object.
  • FIG. 4 is a conceptual diagram illustrating a relationship between a camera coordinate system xyz coordinate system, a x 1 -y 1- z 1 coordinate system transformed such that a straight line connecting the origin of the camera coordinate system and the ellipse is the z 1 axis.
  • FIG. 5 is a conceptual diagram showing the relationship between x 1 -y 1, -z 1 and the coordinate system x 1 -y 1 -z x 2 -y 2 -z 2 coordinates the first coordinate system, a rotational transformation by ⁇ .
  • FIG. 6 is an exemplary diagram of a circle marker shown in a camera image.
  • FIG. 7 is an exemplary view of a circle marker viewed from two different viewpoints in a spherical object drawn so that two circle markers cross 90 degrees.
  • FIG. 8 is a conceptual diagram illustrating a relationship between a circle and a stereo camera image in a three-dimensional space.
  • FIG. 9 is a conceptual diagram illustrating restoring a circle by crossing two elliptical cones.
  • a circle 110 in a three-dimensional space is projected as an ellipse 210 on a two-dimensional camera image surface 200, and the shape of the ellipse depends on the relative posture (position and direction) of the camera and the circle.
  • the circle 110 is placed perpendicular to the camera center line at a distance away from the projection point 220 of the camera along the camera center line 221, the circle shown in the camera image is displayed.
  • the shape looks exactly as a circle 211, but if the circle plane 100 is distorted at an angle with the camera center line 221 as shown in Fig. 2 (b), the circle 110 is shown as an ellipse 212 in the camera image. do.
  • the ellipse shown in the image becomes shorter and shorter as the direction of the circle is distorted with the camera center line.
  • the image is shown as one straight line 213.
  • the position and shape of the ellipse projected on the camera image plane are different depending on the position of the circle center in the three-dimensional space.
  • the equation of the circle in three-dimensional space can be obtained from the position and shape of the ellipse, which is the image of the spherical object seen on the camera image plane, the center position of the circle and the direction of the circle (normal of the plane where the circle is located) Direction) can be calculated.
  • a circle marker is drawn on the surface of a spherical object, and an equation of a circle marker in three-dimensional space is calculated from an ellipse that is a shape of a circle marker shown in an image at an arbitrary time point.
  • the flight speed and the rotation speed of the spherical object are calculated from the center point position and direction of the circle marker calculated at two time points.
  • Step-1 Extract the circle marker image point from the acquired camera image.
  • Process-2 Derivation of the ellipse from the circle marker image point.
  • Step-4 Calculation of the position of the center point of the circle marker and the direction of the circle marker from the equation of the circle marker
  • Process-5 Calculation of the flying speed and the rotational speed of the spherical object using the position of the center point of the circle marker and the direction of the circle marker
  • Process-3 and Process-4 are not clearly distinguished in the following description, but are divided for the purpose of conceptual understanding.
  • the acquired image is a mono image and a stereo image
  • the image is acquired.
  • the fundamental principle of the method of estimating the flight and rotational speeds of a spherical object from the basic principle, that is, the basic principle of calculating the equation of the circle marker from the ellipse shown in the image is the same, but the specific calculation method is different.
  • the acquired image is a mono image and a case where the stereo image is divided will be described.
  • the sphere in order to remove this limitation, the sphere must have two or more circle markers and the circle markers must not be parallel.
  • the position and size of the circle marker are not limited, the centers of all the circle markers on the surface of the sphere object should coincide with the center of the sphere object for convenience of calculation during the estimation process of the sphere object flight information. In other words, all circle markers should be in a great circle.
  • FIG. 3 is an example of a circle marker 310 marked on a spherical object 300 according to the present invention. It should be noted that when there are several circle markers, the range of possible rotation speed estimation is determined by the number and arrangement of circle markers. For example, when three circle markers are arranged as shown in FIG. 3 (b), when the spherical object 300 rotates by 60 degrees or more between two viewpoints, the rotation speed cannot be calculated, and three circle markers are illustrated in FIG. 3 (c). In this case, if the spherical object rotates more than 90 degrees between two time points, the rotation speed cannot be calculated.
  • the interval between the image acquisition time points should be determined not to exceed the maximum rotation angle that can be calculated using the circle marker when the spherical object 300 rotates at the maximum rotation speed.
  • the maximum rotational speed of the golf ball is 10,000rpm
  • the maximum rotation angle of the golf ball is 60 degrees when the image acquisition time interval is 1msec. Therefore, in consideration of the maximum rotation angle, the number of circle markers is preferably not more than three.
  • the method for estimating flight information of a spherical object from a mono image of a flying spherical object provided by the present invention is a mono image of a flying spherical object acquired at two or more time points, regardless of which method a mono image of a flying spherical object is obtained. It is possible to estimate flight information of spherical objects if they are present.
  • C is a matrix representing an ellipse. It is assumed here that the elliptic matrix C shown in the image is known, and a method of obtaining the elliptic matrix C from the circle marker will be described later.
  • an ellipse cone 230 is formed to connect the camera coordinate system origin 220 and the ellipse 210 of the image plane.
  • K is the camera characteristic matrix and x is the position vector of the points on the circumference, described in camera coordinates.
  • the camera characteristic matrix K is obtained through the camera calibration process.
  • Equation 2 is a general elliptical cone z-axis does not coincide with the central axis of the elliptical cone, so that the camera coordinate system as shown in FIG. Rotate to create a new x 1 -y 1 -z 1 coordinate system.
  • the z 1 axis is toward the center of the ellipse, and when viewed from the camera coordinate system origin in the z 1 axis direction, the short axis of the ellipse coincides with the x 1 axis and the long axis coincides with the y 1 axis.
  • the coordinate conversion equation for converting the elliptical cone into the standard form is as shown in Equation 3 below, and the rotation matrix R 1 is calculated through eigen analysis of K T CK .
  • ⁇ is calculated through the eigenvalue analysis of the matrix K T CK as in the rotation matrix R 1 .
  • the ⁇ may have a positive value and a negative value, and the sign of the ⁇ value is determined according to the portion of the circle marker. The method of determining the sign of the? value will be described later.
  • Equation 5 Equation 5
  • Equation 5 are intermediate variables calculated in the process of converting an elliptic cone in the x 1 -y 1 -z 1 coordinate system into a cone in the x 2 -y 2 -z 2 coordinate system.
  • the equation of the circle is a circle whose center is ( ⁇ ta, 0, a) in the x 2- y 2- z 2 coordinate system and the radius is r.
  • the sign of the x 2 coordinate value ⁇ ta of the circle center is determined by the sign of ⁇ calculated previously. That is, if the value of phi is positive, the center point of the circle becomes (-ta, 0, a). If the value of phi is negative, the center point of the circle becomes (+ ta, 0, a). If you know the radius r of the circle in advance, you can find a as
  • the circle When the circle is drawn on the plane, the entire shape of the circle, that is, the shape of the ellipse, is shown on the camera image.
  • the circle when the circle is drawn on the spherical object surface like a circle marker, only a part of the ellipse is shown on the camera image, as shown in FIG.
  • the circle marker drawn on the spherical object can be used to determine the sign of the value of ⁇ by using only part of the circle marker. That is, FIG.
  • this plane of the shorter ellipse 320 positive case part (1 + x-axis part) shown is a circle plane candidates in Fig 5-2 420 circle marker in the value of ⁇ is a positive number
  • Figure 6 (b) is negative, as shown partial circle candidate -1 plane 410 in Figure 5 in the case shown the (-x part 1-axis) of the ellipse speed 330, so in the plane of the circle marker The value of phi becomes negative.
  • a mono image at least two viewpoints is required. For each of the mono images acquired from two or more viewpoints, the position of the center point of the circle marker in the three-dimensional space and the normal direction of the circle marker with respect to the camera are calculated through the following process.
  • Step-1 Extract the original marker image points from the image.
  • image points corresponding to the original markers are first extracted from the acquired image.
  • a method of extracting image points having a certain feature is proposed in the field of image processing, one of them is used.
  • the brightness of the color of the circle marker must be contrasted with the brightness of the color of the spherical object. For example, if the sphere is white, the color of the circle marker should be black.
  • Step-2 Obtain the equation of the ellipse using the circle marker image points.
  • an elliptic equation is obtained through a hough transform that is frequently used in the image processing field.
  • Huff transform method for elliptic expression is applicable to partially visible ellipse, so that even if the original marker is partially erased, the result is highly reliable.
  • the Hough transform method can be applied even when there are several ellipses in the image, so that even if there are several circle markers on the spherical object, the elliptic equations can be calculated.
  • Step-3 Determine the sign of the rotation angle ⁇ by rotating the x 1 -y 1 -z 1 coordinate system about the y 1 axis according to which part of the ellipse is visible around the ellipse short axis, and calculate in step-2. Equation 8 and Equation 9 are used to calculate the normal vector of the circle marker and the position of the center point of the circle marker with respect to the camera.
  • Step-4 If there are several circle markers, average the positions of the center points of the circle markers calculated for each circle marker.
  • the position vector in the three-dimensional space of the circle marker center point calculated from the image at the image acquisition time point t 1 is referred to as c 1
  • the velocity vector v of the spherical object is calculated by the following equation.
  • the method of calculating the rotational speed of a spherical object depends on the number of circle markers. If there is only one number of circle markers, the circle marker normal in three-dimensional space calculated from the image at time t 1 is n 1 , and the circle marker normal in three-dimensional space calculated from the image at time t 2 . If the vector is n 2 and the unit vector k perpendicular to the plane of the angle ⁇ between two normal vectors n 1 and n 2 and n 1 and n 2 is calculated, the rotational velocity vector ⁇ of the spherical object is It is calculated by the following equation.
  • one circle marker selected from the image at the time t 1 of image acquisition is the same as any of the circle markers in the image at the time t 2 before the spherical object is calculated. It should be distinguished from the original marker.
  • the first method for this purpose is to acquire the image of a flying object using a color camera by differentiating the colors of the original markers in order to distinguish between the same circle markers in the image of the image acquisition time t 1 and the image of the image acquisition time t 2 .
  • the rotational speed calculation method does not distinguish between the same circle markers at the image acquisition time t 1 and the image acquisition time t 2 , and may actually occur by a try-and-error method.
  • This method calculates the spherical object's rotation speed by matching the circle markers.
  • the angle of rotation of the spherical object between the image acquisition time t 1 and the image acquisition time t 2 should not exceed the maximum calculation angle (the maximum rotation angle is determined by the number and arrangement of circle markers).
  • Step-1 N as the first of all the original marker normal vector n 1 (i) in the image capturing time point t 1 and the group G 1, N 2 of the image acquisition all of the original at the time t 2 the marker normal vector n 2 (j) Let them be group G 2 .
  • Step-2 Select any circle marker normal vector n 1 (i) from group G 1 for the first pairing, and select two circle normals n 2 (j) from group G 2 after calculating the angle ⁇ 1 between the vector between when each of ⁇ 1 is calculated does not exceed the maximum angle of rotation as possible to go to the next step, the excess of between a maximum angle of rotation as possible each ⁇ 1 is calculated from the mating of the next re-try Exclude the pairing between n 1 (i) and n 2 (j), which you choose not to do, and perform step-2 again to try another pairing that does not exceed the maximum computed rotation angle.
  • Step-3 Record the circle marker normal vectors n 1 (i) and n 2 (j) paired in Step 2 so that the pairing is complete so that the next pairing does not try the same pairing again.
  • Step-4 Select random n 1 (i) and n 2 (j) from group G 1 and group G 2 for the second pairing (except for unmatching or mating complete) After calculating the angle angle ⁇ 2 , if the angle angle ⁇ 2 does not exceed the maximum calculable angle of rotation, proceed to the next step; if the angle angle ⁇ 2 exceeds the maximum calculable angle of rotation, do not try again at the next mating. After excluding the pairing between the selected n 1 (i) and n 2 (j) , perform step 4 again to try another pairing that does not exceed the maximum computed rotation angle.
  • Step-5 Calculate the rotational velocity vectors ⁇ 1 and ⁇ 2 of the spherical object using Equation 11 for the angles ⁇ 1 and ⁇ 2 between the two circle marker normal vectors calculated in the previous step.
  • the rotational speed vector ⁇ of the spherical object is calculated.
  • ⁇ 1 or ⁇ 2 has a value of 0 in the above-described step-5
  • the spherical object is rotated about the center axis of the circle marker, and the rotation speed is calculated as 0 in the circle marker.
  • the previously described method finds only two feasible matings and calculates the spherical object's rotational speed vector ⁇ . However, if there is another feasible mating, the new spherical object's rotational speed vector ⁇ 3 is calculated from the mating and the Averaged with the speed vector ⁇ .
  • circle marker-1-t 1 (331) and circle marker-2-t 2 (352) ⁇ rotation angle exceeded 90 degrees (retry)
  • the method of estimating the flight information that is, the flight speed and the rotational speed of the spherical object from the mono image of the flying spherical object acquired at two time points has been described. If there are more mono images of the flying spherical object acquired at the third time point, the flight speed and the rotational speed of the spherical object are calculated using the first time image and the third time point image as described above. (Averaging result from the image of the first time point and the image of the second time point) and averaging can improve the reliability of the calculation result.
  • the stereo image of the spherical object in flight is acquired by any method, and the flight information of the spherical object can be estimated only if the stereo images of the spherical object in flight are acquired from two or more viewpoints.
  • the stereo image of the spherical object in flight is acquired by any method, and the flight information of the spherical object can be estimated only if the stereo images of the spherical object in flight are acquired from two or more viewpoints.
  • Q is other matrix size that is the equation of a cone (dimension) that is given by the following equation as 4 ⁇ 4 in the other cone matrix, Q the degree of (rank) must be 3
  • C is an ellipse matrix representing the equation of an ellipse seen on the camera image plane
  • P is a projection matrix of the camera, obtained through a camera calibration process.
  • the projection matrix of the first camera and the second camera P 1, P 2, respectively, C 1, C 2 indicates the oval matrices that appear in the image of the first camera and a second camera, respectively, and Q 1, Q 2 are each An elliptic cone matrix is formed by connecting an ellipse of the coordinate system origin and the image plane of the first camera and the second camera.
  • the elliptical cones Q 1 and Q 2 are elliptical cones generated from one circle, that is, the curves formed by the intersection of the elliptical cones Q 1 and Q 2 become circles. Note that when two elliptical cones Q 1 and Q 2 intersect, two circles are generated as shown in FIG. 9.
  • C ([lambda]) at that time represents an equation of a plane containing a curve generated by crossing two cones. Therefore, by calculating ⁇ such that the rank of the C ( ⁇ ) matrix is 2, the plane equation can be obtained.
  • I 2 , I 3 and I 4 mean the coefficients of the cubic, quadratic and linear terms of the determinant of C ( ⁇ ), respectively.
  • Equation 19 is used to distinguish which one circle marker selected from the first camera image is the same circle marker among the circle markers of the second camera image when the number of circle markers is multiple. Can be. The detailed method is described later.
  • the circle facing the camera can only see the side facing the camera and not the other side, so the first camera and the first The plane viewed from the two cameras is the same plane, that is, the circle plane candidate-1 410 in FIG. 9 becomes the plane of the circle.
  • the equation of the circle can be obtained by solving the above-described equation of the circle plane and the equation of the cone Q 1 .
  • the same result can be obtained by using the cone Q 2 instead of the cone Q 1 and solving the above-described equation of the planar equation and the equation of the cone Q 2 by [Equation 16].
  • Equation 19 is a value calculated from an ellipse of the first camera and the second camera image of the same circle, if the image of the same circle C 1 and C 2 is the ellipse of the image of the first camera and the second camera If not, Equation 19 is not satisfied. Therefore, it is possible to distinguish whether ellipses C 1 and C 2 are images of the same circle depending on whether the equation (19) is satisfied.
  • stereo images from at least two viewpoints are required.
  • the center point position of the circle marker in three-dimensional space and the normal vector of the circle marker for the camera are calculated by the following process.
  • Step-1 Image points of a circle marker are extracted from a first camera image and a second camera image acquired at two or more viewpoints. This process is the same as using a mono image.
  • Step-2 Obtain the equations of the ellipse using the circle marker image points in each image. This process is the same as using a mono image.
  • Step-3 Find the absolute value of ⁇ for all possible cases where several ellipses of the first camera and several ellipses of the second camera calculated in step-2 are calculated, and
  • Step-4 The equations of the plane in which the circle-marker is located are calculated and calculated for the ellipses C 1 (i) in the first camera image and the ellipses C 2 (j) in the second camera image. A plane viewed from the first camera and the second camera is selected among the two planes.
  • Step-5 Calculate the equation of the circle by combining the plane equation and the conical equation [Equation 17] calculated in step-4, and find the position c k in the three-dimensional space of the circle-marker from it, and the number of successful matings. If there are several, average c k . Further, the normal vector n k of the circle marker plane is obtained from the plane equation.
  • the position vector in the three-dimensional space of the circle marker center point calculated from the stereo image at the image acquisition time t 1 is c 1
  • the three-dimensional space of the circle marker center point calculated from the stereo image at the image acquisition point t 2 is calculated. If the position vector at is c 2 , the flight velocity vector v of the spherical object is calculated by Equation 11 as in the case of a single image.
  • the rotational speed vector ⁇ diagram won marker normal direction vector calculated from a stereo image of the circle markers normal direction vector n 1 (k) and the image acquisition time t 2 calculated by the stereo images in the image capturing time t 1 of the spherical object n 2 ( k)
  • the method of calculating the rotational speed of the spherical object by matching the circle markers that can actually occur by the try-and-error method can be calculated. Can be.
  • the first object and the third time point image are used as described above.
  • the reliability of the results can be improved by obtaining the flight speed and rotation speed, and averaging the results from the images of the first and second views.
  • the method of estimating the flight information of the spherical object from the mono image in flight of the spherical object inscribed with the circle marker and the method of estimating the flight information of the spherical object from the stereo image in flight has been described.
  • the method of estimating the flight information of the spherical object in flight using the circle marker of the spherical object provided in the present invention can be applied only to the method using the stereo image when the size of the spherical object is not known in advance. If the size of the object can be known in advance, both a mono image and a stereo image can be used.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Physical Education & Sports Medicine (AREA)
  • Multimedia (AREA)
  • Image Analysis (AREA)
  • Length Measuring Devices By Optical Means (AREA)

Abstract

La présente invention concerne un procédé d'estimation d'informations de vol d'objet sphérique à l'aide de marqueur en cercle, des caractéristiques géométriques du marqueur en cercle marquées sur l'objet sphérique étant utilisées pour estimer les informations de vol de l'objet sphérique, le procédé comprenant les étapes consistant à : extraire des points image du marqueur en cercle à partir d'images de caméra provenant de différents points de vue ; obtenir des expressions ovales provenant des différents points de vue à partir des points image extraits du marqueur en cercle ; et calculer une vitesse de vol et une vitesse de rotation de l'objet sphérique à partir d'un vecteur de position et d'un vecteur de ligne normale de chacune des expressions ovales. A cette fin, selon le procédé d'estimation d'informations de vol d'objet sphérique à l'aide de marqueur en cercle de la présente invention, le marqueur en cercle peut être identifié indépendamment d'une couleur ou brillance et similaire du marqueur en cercle par extraction d'une expression ovale de points image du marqueur en cercle à partir d'une image monoscopique ou d'une image stéréoscopique, conversion de l'expression ovale à travers une relation géométrique et analyse du nombre de cas que les points image du marqueur en cercle peuvent avoir.
PCT/KR2012/009747 2011-12-01 2012-11-16 Procédé d'estimation d'informations de vol d'objet sphérique à l'aide de marqueur en cercle WO2013081322A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
KR10-2011-0127950 2011-12-01
KR1020110127950A KR101246975B1 (ko) 2011-12-01 2011-12-01 원 마커를 이용한 구형물체 비행정보 추정 방법

Publications (1)

Publication Number Publication Date
WO2013081322A1 true WO2013081322A1 (fr) 2013-06-06

Family

ID=48535726

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/KR2012/009747 WO2013081322A1 (fr) 2011-12-01 2012-11-16 Procédé d'estimation d'informations de vol d'objet sphérique à l'aide de marqueur en cercle

Country Status (2)

Country Link
KR (1) KR101246975B1 (fr)
WO (1) WO2013081322A1 (fr)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103630117A (zh) * 2013-11-01 2014-03-12 北京邮电大学 用于视觉方法测量高尔夫球运动参数的球面标记
CN106643662A (zh) * 2016-09-20 2017-05-10 深圳市衡泰信科技有限公司 球体及其高速旋转运动参数检测方法
EP3309741A4 (fr) * 2015-06-12 2018-12-12 Golfzon Co., Ltd. Procédé et dispositif de détection de balle en mouvement

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR100871595B1 (ko) * 2007-10-09 2008-12-02 박선의 고속카메라를 이용한 구형물체의 비행정보 측정 시스템
KR100937922B1 (ko) * 2009-02-12 2010-01-21 엔지비스 주식회사 구형 물체의 비행정보 측정 시스템 및 방법
KR20100092357A (ko) * 2009-09-03 2010-08-20 엔지비스 주식회사 구형 물체
KR101044887B1 (ko) * 2009-06-23 2011-06-28 동서대학교산학협력단 고속 씨씨디 카메라의 촬상 이미지를 이용한 골프공의 비행데이터 측정방법

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR100871595B1 (ko) * 2007-10-09 2008-12-02 박선의 고속카메라를 이용한 구형물체의 비행정보 측정 시스템
KR100937922B1 (ko) * 2009-02-12 2010-01-21 엔지비스 주식회사 구형 물체의 비행정보 측정 시스템 및 방법
KR101044887B1 (ko) * 2009-06-23 2011-06-28 동서대학교산학협력단 고속 씨씨디 카메라의 촬상 이미지를 이용한 골프공의 비행데이터 측정방법
KR20100092357A (ko) * 2009-09-03 2010-08-20 엔지비스 주식회사 구형 물체

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103630117A (zh) * 2013-11-01 2014-03-12 北京邮电大学 用于视觉方法测量高尔夫球运动参数的球面标记
EP3309741A4 (fr) * 2015-06-12 2018-12-12 Golfzon Co., Ltd. Procédé et dispositif de détection de balle en mouvement
CN106643662A (zh) * 2016-09-20 2017-05-10 深圳市衡泰信科技有限公司 球体及其高速旋转运动参数检测方法

Also Published As

Publication number Publication date
KR101246975B1 (ko) 2014-02-27

Similar Documents

Publication Publication Date Title
WO2015194864A1 (fr) Dispositif de mise à jour de carte de robot mobile et procédé associé
WO2015194867A1 (fr) Dispositif de reconnaissance de position de robot mobile utilisant le suivi direct, et son procédé
WO2018199701A1 (fr) Procédé de fourniture de contenu et appareil associé
WO2014109545A1 (fr) Appareil et procédé de détection d'une balle en mouvement
WO2016200197A1 (fr) Procédé et appareil de détection de geste dans un système de coordonnées dans l'espace basé sur un utilisateur
WO2013081322A1 (fr) Procédé d'estimation d'informations de vol d'objet sphérique à l'aide de marqueur en cercle
WO2016200208A1 (fr) Procédé et dispositif de détection de balle en mouvement
WO2014109546A1 (fr) Appareil et procédé de détection d'une balle en mouvement
WO2020101420A1 (fr) Procédé et appareil de mesurer des caractéristiques optiques d'un dispositif de réalité augmentée
WO2020159076A1 (fr) Dispositif et procédé d'estimation d'emplacement de point de repère, et support d'enregistrement lisible par ordinateur stockant un programme informatique programmé pour mettre en œuvre le procédé
WO2017007166A1 (fr) Procédé et dispositif de génération d'image projetée et procédé de mappage de pixels d'image et de valeurs de profondeur
WO2020235734A1 (fr) Procédé destiné à estimer la distance à un véhicule autonome et sa position au moyen d'une caméra monoscopique
JP3626400B2 (ja) ターゲットマーカを用いた対象物の位置計測方法及びロボットシステム
JP2003065714A (ja) カメラ・キャリブレーションのためのガイド装置及びガイド方法、並びに、カメラ・キャリブレーション装置
JP2559939B2 (ja) ３次元情報入力装置
WO2012144848A2 (fr) Procédés de reconnaissance tridimensionnelle par l'acquisition d'une image bidimensionnelle constituée de multiples faisceaux de lumière transmise présentant un angle de grossissement préétabli
WO2019135462A1 (fr) Procédé et appareil d'étalonnage entre un dispositif de reconnaissance de mouvement et un visiocasque utilisant un algorithme de réglage de faisceau
WO2023063661A1 (fr) Procédé de génération d'ensemble d'apprentissage, appareil de génération d'ensemble d'apprentissage et système de génération d'ensemble d'apprentissage
WO2022014805A1 (fr) Procédé de construction de plate-forme mobile d'espace virtuel par le biais d'une estimation de coordonnées 3d par covariance croisée
WO2022075691A1 (fr) Appareil et procédé de détection de mouvement de sphère se déplaçant sur un plan, à l'aide d'une caméra, et appareil et procédé de détection de mouvement de balle de golf se déplaçant sur un tapis de putting
WO2021020883A1 (fr) Dispositif et procédé de balayage tridimensionnel
WO2022045819A1 (fr) Dispositif électronique permettant de fournir un contenu de réalité augmentée et son procédé de fonctionnement
WO2017003257A1 (fr) Dispositif et procédé permettant de reconnaître un état de surface de route
WO2022211377A1 (fr) Procédé de calcul de rotation pour balle de golf se déplaçant après avoir été frappée et appareil de calcul de rotation à l'aide dudit procédé
WO2020149527A1 (fr) Appareil et procédé de codage dans un système d'appareil photo de profondeur structuré

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 12854396

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 12854396

Country of ref document: EP

Kind code of ref document: A1