Parameter calibration method, device and system based on microlens array visual identification

Abstract

The present invention provides a parameter calibration method, device and system based on microlens array visual identification. The method includes: taking multiple pictures of the visual identification at different turntable rotation angles by a camera; determining the projection position of the viewpoint of the picture according to the turntable rotation angle corresponding to each picture, and detecting the center position of the moiré fringe of each picture, and determining the corresponding straight line equation according to the projection position of the viewpoint corresponding to each picture and the center position of the moiré fringe; using the LOF outlier detection algorithm to detect whether there are outliers in the intersection of the straight line equation, if so, the outliers are removed, and the initial position of the moiré fringe is determined by the centroid method, if not, the initial position of the moiré fringe is determined by the least squares method; determining the moiré fringe movement distance and line of sight angle of each picture, and determining the slope according to the moiré fringe movement distance and line of sight angle. The technical solution described in the present invention is conducive to improving the position and posture measurement accuracy of the microlens array visual identification.

Description

Technical Field

[0001] The present invention relates to the field of microlens array vision technology, and in particular to a parameter calibration method, device and system based on microlens array vision identification. Background Art

[0002] In the field of vision-based posture measurement, the posture measurement method based on microlens array visual identification can achieve higher measurement accuracy and has a wide range of application backgrounds in robot positioning, aerospace and other fields.

[0003] As key parameters in posture measurement, the initial position and slope of the moiré fringes directly affect the accuracy of posture measurement. However, due to the limitations of the processing technology, the initial position of the moiré fringes will inevitably shift during the processing of microlens array visual identification, which will greatly increase the measurement error. In addition, the line of sight angle and the moving distance of the moiré fringes are linearly related, but the slope of the linear relationship is unknown and changes with the changes in the processing technology. Summary of the invention

[0004] The problem solved by the present invention is how to improve the position and posture measurement accuracy of microlens array visual identification.

[0005] To solve the above problems, the present invention provides a parameter calibration method based on a microlens array visual marker, comprising: taking a plurality of pictures of the visual marker at different turntable rotation angles by a camera, wherein the visual marker is fixed on the turntable and a fixed distance is maintained between the camera and the turntable; determining the projection position of the viewpoint of the picture according to the turntable rotation angle corresponding to each picture, detecting the center position of the moiré fringe of each picture, and determining the corresponding straight line equation according to the projection position of the viewpoint corresponding to each picture and the center position of the moiré fringe; using a LOF outlier detection algorithm to detect whether there are outliers in the intersection of the straight line equation, if so, removing the outliers, and using a centroid method to determine the initial position of the moiré fringe, if not, using a least squares method to determine the initial position of the moiré fringe; determining the moiré fringe movement distance and the line of sight angle of each picture, and determining the slope according to the moiré fringe movement distance and the line of sight angle.

[0006] Optionally, the use of the LOF outlier detection algorithm to detect whether there are outliers in the intersection of the straight line equation includes: determining the kth reachable distance of each point in the kth distance neighborhood of the intersection, determining the local kth local reachable density of the intersection based on the kth reachable distance, determining the kth local outlier factor of the intersection based on the local kth local reachable density, and judging whether the intersection is the outlier based on the size of the kth local outlier factor and a preset threshold.

[0007] Optionally, the determining the initial position of the moiré fringes by using the centroid method includes: determining the initial position of the moiré fringes according to an average value of coordinates of each of the intersection points.

[0008] Optionally, the using of the least square method to determine the initial position of the moiré fringes includes: determining a least square expression according to the distance from each of the intersection points to a straight line, and when the least square expression is minimum, determining the initial position of the moiré fringes according to the corresponding intersection coordinates.

[0009] Optionally, determining the slope according to the moire fringe movement distance and the sight angle includes: determining the slope by fitting according to multiple groups of parameters consisting of the moire fringe movement distance and the sight angle.

[0010] Optionally, the angle range of the turntable includes: ±180° in the vertical direction and ±15° in the pitch direction.

[0011] Optionally, the k is 5 and the preset threshold is 1.5.

[0012] The parameter calibration method based on microlens array visual identification described in the present invention is beneficial to improving the position measurement accuracy of microlens array visual identification by accurately calibrating the initial position and slope of the moiré fringe; the calibration error caused by inaccurate detection of the center position of the moiré fringe is avoided by detecting outliers, and the accuracy of the calibration result of the initial position of the moiré fringe is greatly improved; the initial position of the moiré fringe required for calibration is obtained by using the least squares method, which greatly improves the calculation efficiency and measurement accuracy.

[0013] The present invention also provides a parameter calibration device based on a microlens array visual identification, comprising: a shooting module, used to take multiple pictures of the visual identification at different turntable rotation angles through a camera, wherein the visual identification is fixed on the turntable and a fixed distance is maintained between the camera and the turntable; a processing module, used to determine the projection position of the viewpoint of the picture according to the turntable rotation angle corresponding to each picture, and detect the center position of the moiré fringe of each picture, and determine the corresponding straight line equation according to the projection position of the viewpoint corresponding to each picture and the center position of the moiré fringe; a first calibration module, used to use a LOF outlier detection algorithm to detect whether there are outliers in the intersection of the straight line equation, if so, remove the outliers, and use the centroid method to determine the initial position of the moiré fringe, if not, use the least squares method to determine the initial position of the moiré fringe; a second calibration module, used to determine the moiré fringe movement distance and the line of sight angle of each picture, and determine the slope according to the moiré fringe movement distance and the line of sight angle. The parameter calibration device based on microlens array visual identification described in the present invention has the same advantages as the parameter calibration method based on microlens array visual identification described above over the prior art, which will not be described in detail here.

[0014] The present invention also provides a parameter calibration system based on microlens array visual identification, comprising a computer-readable storage medium storing a computer program and a processor, wherein when the computer program is read and executed by the processor, the parameter calibration method based on microlens array visual identification as described above is implemented. The parameter calibration system based on microlens array visual identification described in the present invention has the same advantages as the parameter calibration method based on microlens array visual identification as described above over the prior art, which will not be described in detail here.

[0015] The present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is read and executed by a processor, the parameter calibration method based on microlens array visual identification as described above is implemented. The advantages of the parameter calibration method based on microlens array visual identification as described above in the computer-readable storage medium of the present invention are the same as those of the prior art, which will not be described in detail here. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a flow chart of a parameter calibration method based on microlens array visual identification according to an embodiment of the present invention;

[0017] Figure 2 A schematic diagram of a picture shooting process according to an embodiment of the present invention;

[0018] Figure 3 A schematic diagram of the structure of a visual identifier according to an embodiment of the present invention;

[0019] Figure 4 This is a schematic diagram of the placement of an industrial camera and a high-precision turntable according to an embodiment of the present invention;

[0020] Figure 5 A schematic diagram of determining the equation of a straight line passing through the moiré fringe position and the viewpoint projection position of each picture according to an embodiment of the present invention;

[0021] Figure 6 A schematic diagram of removing outliers from intersections according to an embodiment of the present invention;

[0022] Figure 7 Schematic diagram of the linear relationship between the sight angle and the moving distance of the moire fringe according to an embodiment of the present invention. DETAILED DESCRIPTION

[0023] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0024] like Figure 1As shown, an embodiment of the present invention provides a parameter calibration method based on a microlens array visual marker, comprising: taking multiple pictures of the visual marker at different turntable rotation angles by a camera, wherein the visual marker is fixed on the turntable and a fixed distance is maintained between the camera and the turntable; determining the projection position of the viewpoint of each picture according to the turntable rotation angle corresponding to the picture, detecting the center position of the moiré fringe of each picture, and determining the corresponding straight line equation according to the projection position of the viewpoint corresponding to each picture and the center position of the moiré fringe; using the LOF outlier detection algorithm to detect whether there are outliers in the intersection of the straight line equation, if so, removing the outliers, and using the centroid method to determine the initial position of the moiré fringe, if not, using the least squares method to determine the initial position of the moiré fringe; determining the moiré fringe movement distance and the line of sight angle of each picture, and determining the slope according to the moiré fringe movement distance and the line of sight angle.

[0025] Specifically, the parameter calibration method based on microlens array visual identification includes:

[0026] (1) Combination Figure 2 and Figure 3 As shown in the figure, the visual marker is fixed on the high-precision turntable, the camera is fixed so that the distance between it and the high-precision turntable is fixed, and the front view is taken. The high-precision turntable is specifically a device that can provide a rotation angle of ±180° in the vertical direction and a pitch angle of ±15° in the pitch direction. In terms of accuracy, the repeatability error of the rotation angle of the two-axis turntable does not exceed 0.01°. The relative position of the camera and the high-precision turntable is shown in the figure. Figure 4 By adjusting the rotation angle of the turntable, a professional industrial camera is used to take multiple pictures of the logo at different angles until the number of pictures reaches the required number.

[0027] (2) Calculate the projection position of the viewpoint of each image taken in step 1 according to the rotation angle of the turntable corresponding to the image, and use image processing methods to detect the center position of the moiré fringe of the image. Select a model of points and direction vectors to describe the straight line, and calculate the points and direction vectors in the straight line model according to the projection position of the viewpoint and the center position of the moiré fringe. Figure 5 As shown, each picture taken can determine a straight line, and multiple pictures can obtain multiple straight lines.

[0028] (3) Calculate the intersection points of all the straight line equations obtained in step 2, and use the LOF outlier detection algorithm to detect whether there are outliers in the intersection points. By detecting outliers, the calibration error caused by inaccurate detection of the center position of the moiré fringe is avoided, and the accuracy of the initial position calibration result of the moiré fringe is greatly improved.

[0029] Let point p be a point in the intersection point set of the line equation. The kth neighborhood of point p is denoted by N k(p). Calculate the kth reachable distance of each point in the kth distance neighborhood of each point:

[0030] reach-dist k (O,p)=max{d k (O),d(O,p)}

[0031] Among them, d k (O) is the kth reachable distance of point O, and d(O,p) is the distance from point O to point p.

[0032] On this basis, the local kth local reachable density of each point is calculated:

[0033]

[0034] Among them, |N k (p)| represents the number of all points in the neighborhood.

[0035] On this basis, the kth local outlier factor of each point is calculated:

[0036]

[0037] LOF k Points with (p) values ​​greater than thresh are outliers.

[0038] In the above process, k and thresh are empirical parameters and can be adjusted according to actual conditions. The parameters recommended by the present invention are: k=5, thresh=1.5.

[0039] (4) Determine whether an outlier is detected in step 3. Figure 6 As shown, if outliers are detected, they are removed and the initial position of the moiré fringe required for calibration is obtained using the centroid method.

[0040] LOF k (p) Points with values ​​greater than thresh are removed. Let the coordinates of the intersection after removing the outliers be (x i ,y i ), then the coordinates of the initial position of the moiré fringe are calculated by the following formula:

[0041]

[0042] Among them, (x, y) is the coordinate of the initial position of the moiré fringe.

[0043] (5) Determine whether an outlier point is detected in step 3. If no outlier point is detected, the least square method is used to obtain the initial position of the moiré fringe required for calibration. The least square method is used to obtain the initial position of the moiré fringe required for calibration. The equation used is a linear equation system, which can avoid the problem that nonlinear optimization methods such as the Gauss-Newton method and the LM method may fall into local minima, greatly improving the calculation efficiency and measurement accuracy.

[0044] First consider a point p to a straight line (a i is a point on the line, n i is the direction vector on the line):

[0045]

[0046] By generalizing from one straight line to multiple straight lines, we get the least squares expression:

[0047]

[0048] The point p where this formula is the smallest is the optimal point. Therefore, taking the derivative of the above formula and taking the extreme value, we get:

[0049]

[0050]

[0051] The final equation is:

[0052]

[0053] Solving this equation, the point p obtained is the final initial position of the moiré fringe.

[0054] (6) Calculate the moiré fringe movement distance and sight angle of each image taken in step 1 to obtain the slope required for calibration.

[0055] Let the initial position of the moiré fringe be (x c0 ,y c0 ), the real coordinates of the viewpoint in the world coordinate system are (x, y, z), and the coordinates of the center of the moiré fringe are (x m ,y m ). Based on this, the sight angle can be solved:

[0056]

[0057] The moving distance of the moiré fringes is:

[0058]

[0059] A set of (θ, s) parameters can be obtained from one calibration photo. Usually, 5-8 sets of (θ, s) parameters can be obtained from 5-8 calibration photos. Figure 7 As shown, the slope of the straight line can be fitted based on these parameters.

[0060] The present invention solves the problem of visual identification parameter calibration in posture measurement based on microlens array visual identification, can accurately calibrate the initial position and slope of the Moire fringe, and provides a new idea for improving the measurement accuracy of posture measurement technology based on microlens array visual identification.

[0061] In this embodiment, by accurately calibrating the initial position and slope of the moiré fringes, it is beneficial to improve the position measurement accuracy of the microlens array visual identification; by detecting outliers, the calibration error caused by inaccurate detection of the center position of the moiré fringes is avoided, and the accuracy of the calibration result of the initial position of the moiré fringes is greatly improved; the least squares method is used to obtain the initial position of the moiré fringes that need to be calibrated, which greatly improves the calculation efficiency and measurement accuracy.

[0062] Optionally, the use of the LOF outlier detection algorithm to detect whether there are outliers in the intersection of the straight line equation includes: determining the kth reachable distance of each point in the kth distance neighborhood of the intersection, determining the local kth local reachable density of the intersection based on the kth reachable distance, determining the kth local outlier factor of the intersection based on the local kth local reachable density, and judging whether the intersection is the outlier based on the size of the kth local outlier factor and a preset threshold.

[0063] Optionally, the determining the initial position of the moiré fringes by using the centroid method includes: determining the initial position of the moiré fringes according to an average value of coordinates of each of the intersection points.

[0064] Optionally, the using of the least square method to determine the initial position of the moiré fringes includes: determining a least square expression according to the distance from each of the intersection points to a straight line, and when the least square expression is minimum, determining the initial position of the moiré fringes according to the corresponding intersection coordinates.

[0065] Optionally, determining the slope according to the moire fringe movement distance and the sight angle includes: determining the slope by fitting according to multiple groups of parameters consisting of the moire fringe movement distance and the sight angle.

[0066] Optionally, the angle range of the turntable includes: ±180° in the vertical direction and ±15° in the pitch direction.

[0067] Optionally, the k is 5 and the preset threshold is 1.5.

[0068] Another embodiment of the present invention provides a parameter calibration device based on a microlens array visual identification, comprising: a shooting module, used to take multiple pictures of the visual identification at different turntable rotation angles through a camera, wherein the visual identification is fixed on the turntable and a fixed distance is maintained between the camera and the turntable; a processing module, used to determine the projection position of the viewpoint of the picture according to the turntable rotation angle corresponding to each picture, and detect the center position of the moiré fringe of each picture, and determine the corresponding straight line equation according to the projection position of the viewpoint corresponding to each picture and the center position of the moiré fringe; a first calibration module, used to use a LOF outlier detection algorithm to detect whether there are outliers in the intersection of the straight line equation, if so, remove the outliers, and use the centroid method to determine the initial position of the moiré fringe, if not, use the least squares method to determine the initial position of the moiré fringe; a second calibration module, used to determine the moiré fringe movement distance and the line of sight angle of each picture, and determine the slope according to the moiré fringe movement distance and the line of sight angle.

[0069] Another embodiment of the present invention provides a parameter calibration system based on microlens array visual identification, including a computer-readable storage medium storing a computer program and a processor. When the computer program is read and executed by the processor, the above-mentioned parameter calibration method based on microlens array visual identification is implemented.

[0070] Another embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is read and executed by a processor, the parameter calibration method based on microlens array visual identification as described above is implemented.

[0071] Although the present invention is disclosed as above, the protection scope of the present invention is not limited thereto. Those skilled in the art may make various changes and modifications without departing from the spirit and scope of the present invention, and these changes and modifications will fall within the protection scope of the present invention.

Claims

1. A parameter calibration method based on microlens array visual identification, characterized in that: include: Taking a plurality of pictures of the visual mark at different rotation angles of the turntable by a camera, wherein the visual mark is fixed on the turntable and a fixed distance is maintained between the camera and the turntable; Determine the projection position of the viewpoint of each picture according to the rotation angle of the turntable corresponding to the picture, detect the center position of the moiré fringes of each picture, and determine the corresponding straight line equation according to the projection position of the viewpoint corresponding to each picture and the center position of the moiré fringes; The LOF outlier detection algorithm is used to detect whether there are outliers in the intersection of the straight line equation. If so, the outliers are removed and the initial position of the moiré fringe is determined by the centroid method. If not, the initial position of the moiré fringe is determined by the least square method; Determine the moire fringe movement distance and the sight angle of each image, and determine the slope according to the moire fringe movement distance and the sight angle; Among them, the use of the LOF outlier detection algorithm to detect whether there are outliers in the intersection of the straight line equation includes: determining the kth reachable distance of each point in the kth distance neighborhood of the intersection, determining the local kth local reachable density of the intersection according to the kth reachable distance, determining the kth local outlier factor of the intersection according to the local kth local reachable density, and judging whether the intersection is the outlier according to the size of the kth local outlier factor and a preset threshold; Wherein, the use of the least square method to determine the initial position of the moiré fringe includes: determining a least square expression according to the distance from each intersection point to a straight line, and when the least square expression is minimum, determining the initial position of the moiré fringe according to the corresponding intersection point coordinates.

2. The parameter calibration method based on microlens array visual identification according to claim 1, characterized in that: The method of using the centroid method to determine the initial position of the moiré fringes includes: determining the initial position of the moiré fringes according to the average value of the coordinates of each of the intersection points.

3. The parameter calibration method based on microlens array visual identification according to claim 1, characterized in that: Determining the slope according to the moire fringe movement distance and the sight angle includes: determining the slope by fitting according to multiple groups of parameters consisting of the moire fringe movement distance and the sight angle.

4. The parameter calibration method based on microlens array visual identification according to claim 1, characterized in that: The angle range of the turntable includes: ±180° in the vertical direction and ±15° in the pitch direction.

5. The parameter calibration method based on microlens array visual identification according to claim 1, characterized in that: The k is 5, and the preset threshold is 1.

5.

6. A parameter calibration device based on microlens array visual identification, characterized in that: include: A shooting module, used for shooting a plurality of pictures of the visual mark at different rotation angles of the turntable through a camera, wherein the visual mark is fixed on the turntable, and a fixed distance is maintained between the camera and the turntable; A processing module, used to determine the projection position of the viewpoint of each picture according to the rotation angle of the turntable corresponding to the picture, detect the center position of the moiré fringes of each picture, and determine the corresponding straight line equation according to the projection position of the viewpoint corresponding to each picture and the center position of the moiré fringes; A first calibration module is used to detect whether there are outliers in the intersection of the straight line equations by using the LOF outlier detection algorithm. If yes, the outliers are removed and the initial position of the moiré fringe is determined by using the centroid method. If no, the initial position of the moiré fringe is determined by using the least square method; A second calibration module is used to determine the moiré fringe movement distance and the sight angle of each image, and determine the slope according to the moiré fringe movement distance and the sight angle; Among them, the use of the LOF outlier detection algorithm to detect whether there are outliers in the intersection of the straight line equation includes: determining the kth reachable distance of each point in the kth distance neighborhood of the intersection, determining the local kth local reachable density of the intersection according to the kth reachable distance, determining the kth local outlier factor of the intersection according to the local kth local reachable density, and judging whether the intersection is the outlier according to the size of the kth local outlier factor and a preset threshold; Wherein, the use of the least square method to determine the initial position of the moiré fringe includes: determining a least square expression according to the distance from each intersection point to a straight line, and when the least square expression is minimum, determining the initial position of the moiré fringe according to the corresponding intersection point coordinates.

7. A parameter calibration system based on microlens array visual identification, characterized in that: It comprises a computer-readable storage medium storing a computer program and a processor, and when the computer program is read and executed by the processor, the parameter calibration method based on microlens array visual identification as claimed in any one of claims 1 to 5 is implemented.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is read and executed by a processor, the parameter calibration method based on microlens array visual identification as described in any one of claims 1 to 5 is implemented.

Images