Non-contact ice friction coefficient measuring method and device

Abstract

The invention discloses a non-contact ice friction coefficient measuring method and device, and relates to the tribology testing technology, and the method comprises the following steps: obtaining a motion image of a slide block on an artificial ice rink by using a camera system; performing preprocessing and target detection on the acquired image to obtain coordinate information of the sliding block in a world coordinate system; and based on the coordinate data of the sliding block, fitting and solving the acceleration of the sliding block by adopting primary function expansion in combination with a Tikhonov regularization method, and carrying out inversion in combination with a friction mechanical model and a kinematic model to obtain an ice friction coefficient. Compared with a traditional contact type method, the in-situ high-precision measurement of the ice friction coefficient can be realized under the condition of no sensor attachment or external force interference, the defects of low measurement precision and strong invasiveness in the prior art are effectively overcome, and the method can be applied to the fields of winter road safety detection and ice and snow sports.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of tribology, in particular to a non-contact ice friction coefficient measurement method and device, mainly measuring the parameter of ice friction coefficient. BACKGROUND

[0002] As a key parameter for evaluating the friction characteristics of ice surface, the accurate measurement of ice friction coefficient has important application value in the fields of winter road traffic safety maintenance, airport pavement icing monitoring, and ice and snow sports equipment design.

[0003] Currently, the commonly used ice friction coefficient measurement method mainly relies on contact sensor technology, which directly contacts the ice surface through the sensor to obtain friction force data. However, the contact measurement method has certain limitations, that is, the contact between the sensor and the ice surface may cause damage to the ice surface or change the friction characteristics of the ice surface, thereby affecting the measurement results. In addition, using an inertial measurement unit (IMU) to measure the acceleration of an ice surface object and obtaining the ice friction coefficient through inversion calculation also has certain problems. The IMU signal fluctuates greatly, making it difficult to accurately obtain the acceleration of the ice surface object, and the need to install a sensor on the object increases the invasiveness of the measurement. Influenced by the above factors, the existing method is difficult to provide high-precision, in-situ and non-interfering measurement in actual application, especially under dynamic measurement conditions.

[0004] To solve the above problems, the present application provides a non-contact ice friction coefficient measurement method and device. The present application uses a high-frame-rate camera system to collect the motion images of the slider on the ice surface, and based on the sub-pixel edge detection method in machine vision technology, accurately extracts the position coordinates of the slider in each frame of image. Then, combined with the Coulomb friction physical model and the Tikhonov regularization algorithm, the ice friction coefficient is obtained based on the inversion of the slider acceleration. This method can realize non-contact, in-situ and high-precision ice friction coefficient measurement, overcoming the problems of low precision, strong invasiveness and difficulty in measuring under dynamic conditions in the prior art. SUMMARY

[0005] The present application provides a non-contact ice friction coefficient measurement method and device based on a camera system and a computer system. By collecting continuous motion images of the slider on the ice surface and using a sub-pixel edge detection method, the position coordinates of the slider in each frame of image are accurately extracted. Combined with the motion model of the slider and the friction mechanics model, the ice friction coefficient is calculated using an inversion algorithm, thereby realizing high-precision in-situ measurement and providing a new solution for non-contact acquisition of ice friction coefficient.

[0006] The measurement device used in the present application includes a camera system and a computer system, and the object to be measured is the ice surface and the moving slider, characterized in that:

[0007] The camera system comprises a high-frame-rate camera, a camera support frame, a level, and a camera calibration board, and the main feature is that it can collect real-time images of the slider on the ice surface without physical contact, and can obtain continuous clear images with a resolution of not less than 960*540 and a frame rate of not less than 100 fps, so as to ensure high-precision measurement under dynamic conditions.

[0008] The computer system comprises a computer and a data connection line, and the main feature is that it receives the image data collected by the camera system and runs an ice friction coefficient inversion algorithm, so as to obtain the ice friction coefficient in situ without relying on sensors or external force interference.

[0009] The ice surface to be measured and the moving slider, and the main feature is that the ice surface to be measured has a uniform and consistent surface structure without obvious defects; and the mass distribution of the slider is uniform.

[0010] The technical solution adopted by the present application to solve the technical problem is as follows:

[0011] A non-contact ice friction coefficient measurement method, characterized in that it comprises the following steps:

[0012] Step S1: collecting experimental images of the calibration board and the slider movement on the ice surface;

[0013] Step S2: processing the images and outputting the position coordinates of the slider;

[0014] Step S3: calculating the motion acceleration of the slider;

[0015] Step S4: inversely obtaining the ice friction coefficient based on the relationship between the acceleration and the friction mechanics model;

[0016] The further technical solution is that step S1 comprises the following sub-steps:

[0017] Step S1-1: photographing the calibration board with known geometric dimensions at different positions in the image area to obtain an image sequence of the calibration board;

[0018] Step S1-2: using the camera system to continuously collect images of the slider movement to obtain complete position information of the slider movement.

[0019] Further, when photographing the calibration board, the position distribution of the calibration board in the image area includes the edge of the picture, the center, and covers the typical trajectory area of the slider movement;

[0020] Further, when photographing the moving slider, the surface of the slider is subjected to anti-reflection processing, and identification mark points are set.

[0021] Steps S2 and S3 need to use code editing algorithms to realize.

[0022] Further, step S2 comprises the following sub-steps:

[0023] In step S2-1, the edge features of the calibration board are extracted by using a sub-pixel edge detection method, the conversion relationship between the pixel coordinates and the world coordinates is established after evaluating the image distortion degree;

[0024] In step S2-2, the color slider image obtained by shooting is converted into a gray-scale image, and the image is binarized based on a preset threshold or an adaptive threshold;

[0025] In step S2-3, the frame difference method is used to detect the motion target region of the binarized image, and then the center position of the motion target is determined based on the least square fitting circle method, and a mask region is established in each frame image with the center as the center and in combination with a preset or adaptive radius, and the image is subjected to mask processing;

[0026] In step S2-4, the coordinates of the slider surface marker point contour are extracted based on the sub-pixel edge detection method, and the centroid position thereof is calculated according to the contour;

[0027] In step S2-5, the pixel coordinates of the slider position are converted into physical coordinates in the world coordinate system based on the camera calibration result.

[0028] Further, in step S2-1, the algorithm of sub-pixel edge detection is used to obtain the edge coordinates of the marker points, and the detection method is proposed by Trujillo-Pino et al. in the paper “Accurate Subpixel Edge Location based on Partial Area Effect”;

[0029] Further, in step S2-1, the algorithm of sub-pixel edge detection needs to set a threshold parameter, and the specific formula is

[0030]

[0031] In formula 1, Gs represents the gray value of the pixel point, and Sp represents the area of the pixel point p1. The real contour passes through the pixel point p1, and divides the area into two parts, and the part close to the pixel point p2 is defined as Sp(p2), and Sp(p3) is the same.

[0032] Further, in step S2-1, the results of sub-pixel edge detection of the calibration board are combined, and the total variance of the pixel numbers of the four edges of the calibration board is used as an evaluation index, and the calculation formula of the total variance is

[0033]

[0034] In formula 2, σ 2is the total variance, N represents the total number of data, x i represents the number of pixels of one side of the i-th calibration plate, and μ represents the population mean.

[0035] Further, in step S2-1, in order to establish the relationship between the pixel coordinates and the world coordinates, that is, to determine the relationship between the geometric position of the slider in space and the corresponding point in the image, a geometric model of camera imaging needs to be established, and the parameters thereof are solved. First, the world coordinate system is converted to the camera coordinate system through rigid body transformation, then converted from the camera coordinate system to the image coordinate system through perspective projection, and finally converted from the image coordinate system to the pixel coordinate system through affine transformation. The specific conversion relationship of this process is

[0036] m = K[R t]M = HM Formula 3

[0037] In formula 3, m is the pixel coordinate system coordinate, K is the intrinsic matrix of the camera, [Rt] is the extrinsic matrix of the camera, R is the rotation matrix, t is the translation matrix, M is the coordinate in the world coordinate system, and H is the matrix product of the intrinsic matrix K and the extrinsic matrix [Rt];

[0038] Further, in step S2-1, the formula for converting the world coordinate system to the camera coordinate system is

[0039]

[0040] In formula 4, X c , Y c , and Z c are the coordinates in the camera coordinate system, and X w , Y w , and Z w are the coordinates in the world coordinate system.

[0041] Further, in step S2-1, the formula for converting the camera coordinate system to the image coordinate system is

[0042]

[0043] In formula 5, x and y are the coordinates in the image coordinate system, and f is the focal length.

[0044] Further, in step S2-1, the formula for converting the image coordinate system to the pixel coordinate system is

[0045]

[0046] In formula 6, u and v are the coordinates in the pixel coordinate system, c x and c y are the origin coordinates in the image coordinate system, f x is the number of pixels per millimeter in the x direction of the image.y the number of pixels per millimeter in the y direction of the image;

[0047] Further, in step S2-1, the f x and f y are solved to obtain the coordinate conversion proportionality coefficient between the world coordinate system and the pixel coordinate system.

[0048] Further, in step S2-2, a threshold value needs to be set to distinguish the slider from the background during the binarization process.

[0049] Further, in step S2-3, the present application determines the center position of the moving target by using the least square fitting circle method, and the general equation of the circle is expressed as

[0050] x 2 +y 2 +a1x+a2y+a3=0 formula 7

[0051] In formula 7, x and y represent the horizontal and vertical coordinates of the data points, and a1, a2, and a3 are the parameters to be solved.

[0052] Further, in step S2-3, the least square fitting circle minimizes the sum of squares of distance errors of all data points to the fitting circle, and the error function can be expressed as

[0053]

[0054] In formula 8, E(a1, a2, a3) is the error function, n is the number of data points, x i , and y i are the horizontal and vertical coordinates of the i-th data point, respectively.

[0055] Further, in step S2-3, by minimizing the error function E(a1, a2, a3), the parameters a1, a2, and a3 to be solved are obtained, and the center coordinates and the radius of the fitting circle are determined.

[0056] Further, in step S2-4, a sub-pixel edge detection algorithm is used to detect the contour coordinates of the marker point on the slider in the pixel coordinate system, and the centroid is calculated according to the detected contour coordinates, and the centroid calculation formula is

[0057]

[0058] In formula 9, x c , y c are the horizontal and vertical coordinates of the marker centroid in the pixel coordinate system, respectively, n is the number of contour points, x i , and y iare the horizontal and vertical coordinates of the contour point in the pixel coordinate system, respectively.

[0059] Further, in step S2-5, the centroid coordinates of the marker points are converted into coordinates in the world coordinate system according to the coordinate conversion scale factor obtained in formula 6.

[0060] Step S3 includes the following sub-steps:

[0061] Step S3-1: a kinematic model of the slider is established, and an acceleration equation is reconstructed.

[0062] Step S3-2: the Tikhonov regularization method is used to solve the coefficient term corresponding to the basis function using the identified slider coordinate information, so as to obtain the acceleration equation of the slider.

[0063] Further, in step S3-1, the formula for acceleration reconstruction is

[0064] a(t) = B · R n+1 Formula 10

[0065] Further, in formula 10, B 1×(n+1) is an n+1-dimensional row vector, where each element b(j) = t j-1 (j = 1, 2, …, n+1) represents a basis function term, R (n+1)×1 = (r1, r2, …, r n+1 ) T is a coefficient term corresponding to the basis function, and n is the order of the basis function expansion. At this time, the problem of solving a(t) is converted into solving the coefficient vector R n+1 . Once a set of R n+1 is obtained, the acceleration of the slider motion can be obtained.

[0066] Further, in step S3-2, the Tikhonov regularization algorithm is used to invert the acceleration of the slider, and the calculation formula of the algorithm is

[0067] min{||AR n+3 -W|| 2 +λ 2 ||R n+3 || 2} Formula 11

[0068] In formula 11, ||*|| represents the 2-norm of a vector, W m×1 = (w1, w2, …, w m ) T is an m-dimensional column vector, is the centroid coordinates of the marker points of the slider, and λ is the Tikhonov regularization parameter.

[0069] Step S4 comprises the following sub-steps:

[0070] Step S4-1: establishing a friction mechanics model between the slider and the ice surface;

[0071] Step S4-2: establishing a motion-force relationship of the slider, and performing inversion of the ice friction coefficient;

[0072] Further, in step S4-1, it is assumed that the Coulomb friction model is met between the slider and the ice surface, and the mechanical relationship can be expressed as

[0073] f = mu * N formula 12

[0074] In formula 12, f is the friction force, mu is the ice friction coefficient, and N is the support force of the ice surface on the slider.

[0075] Further, in step S4-2, the relationship between the ice friction coefficient and the acceleration is established according to Newton's second law, and the expression is

[0076] mu = a * g -1 Formula 12

[0077] In formula 12, a is the acceleration of the slider, and g is the gravitational acceleration. Finally, formula 12 realizes the non-contact inversion of the ice friction coefficient.

[0078] Further technical solutions are that when the method is used, the following matters should be noted:

[0079] When recording with the camera system, the aperture of the camera should be adjusted so that the slider in the picture can be better distinguished from the background;

[0080] When identifying the calibration plate and the marker point, the setting of the algorithm parameters should be paid attention to, so that the edge of the identified marker point is not distorted;

[0081] The expansion order of the acceleration reconstruction basis function should be reasonably set to avoid distortion of the result;

[0082] The method can effectively measure the motion information of the slider and invert the friction coefficient of the ice surface, and the quality of the ice surface can be measured. BRIEF DESCRIPTION OF DRAWINGS

[0083] In order to more clearly illustrate the implementation method of the present application and the prior art, the specific implementation method of the present application will be described in detail below with reference to the accompanying drawings.

[0084] Figure 1 It is a non-contact ice friction coefficient measuring device; In the drawings: 1-camera 2-camera support frame 3-level meter 4-Computer 5-Data connection line 6-Moving slider 7-Ice surface to be measured

[0085] Figure 2 A flow chart of a non-contact ice friction coefficient measurement method;

[0086] Figure 3 A detailed description of sub-step S2-1 in a non-contact ice friction coefficient measurement method;

[0087] Figure 4 A detailed description of steps S2 and S3 in a non-contact ice friction coefficient measurement method;

[0088] Figure 5 Identify the ice friction coefficient of the ice surface corresponding to the moving track of the slider and its recognition points. DETAILED DESCRIPTION

[0089] In order to more clearly illustrate the features and advantages of the present application, the following will be a detailed description of the present application through examples. It should be particularly pointed out that the following examples are only examples and are intended to help understand the technical solutions of the present application; unless otherwise specified, all technical terms used in the specification are consistent with the meanings understood by those of ordinary skill in the art.

[0090] The main measuring device used in the present application (its arrangement diagram is shown in Figure 1 The main measuring device used in the present application (its arrangement diagram is shown in The main measuring device used in the present application (its arrangement diagram is shown in

[0091] The main measuring device used in the present application (its arrangement diagram is shown in Figure 2 A non-contact ice friction coefficient measurement method and device (its flow chart is shown in

[0092] Step S1-1: Photograph a calibration plate with known geometric dimensions at different positions in the image area;

[0093] Step S1-2: Record a continuous image sequence of the moving slider on the surface treated by the camera system;

[0094] Step S2-1: Extract the edge features of the calibration board using the sub-pixel edge detection method, evaluate the degree of picture distortion and establish the coordinate conversion relationship (conversion relationship as shown in Figure 3

[0095] Step S2-2: After the image of the slider movement is subjected to grayscale processing, it is subjected to binaryzation processing;

[0096] Step S2-3: The frame difference method is used to detect the moving target area in the image, and the center of the moving target is determined based on the least square method, and then the image is subjected to mask processing;

[0097] Step S2-4: The position coordinates of the slider in the image are determined based on the sub-pixel edge detection method;

[0098] Step S2-5: The coordinates of the slider in the pixel coordinate system are converted into the coordinates in the world coordinate system;

[0099] Step S3-1: According to the kinematic model of the slider, the acceleration equation of the slider is reconstructed;

[0100] Step S3-2: The Tikhonov regularization method is used to solve the slider acceleration equation using the identified slider coordinate information;

[0101] Step S4-1: A friction mechanics model between the slider and the ice surface is established;

[0102] Step S4-2: The motion-force relationship of the slider is established, and the ice friction coefficient is inverted.

[0103] Further, in step S1-1, when the calibration board is photographed, the position distribution of the calibration board in the image area includes the picture edge, the center, and covers the typical trajectory area of the slider movement;

[0104] Further, in step S1-2, when the moving slider is photographed, the slider surface is adhered with black light-absorbing material, and a white reflective marker is provided at the center of rotation of the slider;

[0105] Further, in step S1-2, the focal length, aperture, shooting frame rate and resolution of the selected camera need to be parameterized.

[0106] Further, in step S2-1, the threshold value of the sub-pixel edge detection grayscale recognition needs to be reasonably set;

[0107] Further, in step S2-1, the geometric model of camera imaging needs to be established and the parameters of the geometric model are solved to determine the coordinate conversion proportionality coefficient.

[0108] Further, in step S2-2, the threshold value of the image binaryzation needs to be reasonably set. ​

[0109] Further, in step S2-3, when the image is masked, a reasonable circular area needs to be generated in combination with the specific geometric size of the slider. The new circle is determined according to the centroid coordinates, and the radius should be set according to the actual size of the slider to ensure that the slider can be completely covered and the background can be effectively shielded.

[0110] Further, in step S2-4, the threshold value of the sub-pixel edge detection gray recognition needs to be reasonably set.

[0111] Further, in step S3-1, the order of the basis function of the reconstructed acceleration needs to be determined.

[0112] Further, in step S3-2, the Tikhonov regularization parameter λ is determined by the generalized cross-validation method (GCV).

[0113] Further, in step S4-1, it is assumed that the friction between the slider and the ice surface conforms to the Coulomb friction model.

[0114] Further, in step S4-2, the relationship between the motion and the force of the slider is established according to Newton's second law, and the calculated acceleration is brought into the above relationship to inversely calculate the ice friction coefficient.

[0115] Based on the above technical scheme, the application designs a non-contact ice friction coefficient measurement method and device. The method has high recognition accuracy, and the device is easy to operate, can efficiently measure the ice friction coefficient, and overcomes the problems of low precision and strong invasiveness in the prior art.

[0116] The above is only the preferred embodiment of the application, and it should be pointed out that for ordinary skilled persons in the technical field, several improvements and modifications can be made without departing from the technical principles of the application, and these improvements and modifications should be considered as the protection scope of the application.

Claims

1. A non-contact ice friction coefficient measurement device, comprising a camera system and a computer system, and the measurement object comprises an ice surface to be measured and a moving slider, characterized in that: the camera system comprises a high-frame-rate camera, a camera support frame, a level, and a photographic calibration board, and the main feature is to collect the moving image of the slider on the ice surface in real time without physical contact, and the system can collect clear images with a resolution of not less than 960*540 and a frame rate of not less than 100 fps; the ice surface to be measured and the moving slider are the measurement objects, and the main feature is that the ice surface is an artificially prepared ice field, and the moving slider is the object of friction coefficient measurement; the computer system comprises a computer and a data connection line, and the main feature is to receive the image data of the camera system and run the ice friction coefficient inversion algorithm to obtain the ice friction coefficient in situ without sensor attachment or external force interference; the camera system adjusts the camera posture through a horizontal adjustment device to make the optical axis of the camera lens perpendicular to the ice surface, and adjusts the focal length to capture all positions of the slider in different motion trajectories; the ice surface to be measured is an artificially prepared ice field with a uniform and consistent surface structure without obvious defects; the moving slider is the object of friction coefficient measurement, and the surface of the slider is provided with a light-eliminating material to reduce optical interference, and an identification mark point is arranged at the center of rotation of the slider to facilitate high-precision detection and analysis of the slider position by the computer system; the computer system can position and reconstruct the motion trajectory of the slider based on the image data obtained by the camera system, and use the basis function expansion combined with the Tikhonov regularization method to fit and solve the acceleration of the slider for ice friction coefficient inversion calculation. The measurement method comprises the following steps: (a) Step S1: experimental image collection of the calibration board and the movement of the slider on the ice surface, characterized in that S1 comprises the following sub-steps: (1) Sub-step S1-1: photographing the calibration board with known geometric dimensions at different positions in the image area to obtain an image sequence of the calibration board; 2. A non-contact ice friction coefficient measuring device as described in claim 1, wherein, (2) Sub-step S1-2: as described in claim 3, applying a certain initial speed to the slider on the ice surface which has been treated with surface light elimination and provided with an identification mark point, so that the slider slides within the field of view of the camera system, and the movement process of the slider is recorded by using the camera system to obtain a continuous image sequence of the movement of the slider.

3. A non-contact ice friction coefficient measuring device as described in claim 1, wherein, (b) Step S2: processing the image and outputting the position coordinates of the slider, characterized in that S2 comprises the following sub-steps:

4. A non-contact ice friction coefficient measuring device as described in claim 1, wherein, (1) Sub-step S2-1: using a sub-pixel edge detection method to extract the edge features of the calibration board, evaluating the image distortion degree, and establishing the conversion relationship between the pixel coordinates and the world coordinates; 5. A method of measuring the coefficient of ice friction non-contact, characterized by, (2) Sub-step S2-2: converting the photographed color slider image into a grayscale image, and performing binaryzation processing on the image based on a preset threshold or an adaptive threshold. ​ ​ ​ ​ ​ ​ (3) Sub-step S2-3: frame difference processing is performed on the binary image to detect the moving target region, and then the center position of the moving target is determined based on a least square fitting circle method, and a mask region is established in each frame image with the center as the center and in combination with a preset or adaptive radius, and the image is subjected to mask processing; (4) Sub-step S2-4: the profile of the slider surface marker point is extracted based on a sub-pixel edge detection method, and the centroid position thereof is calculated as the pixel coordinates of the slider in the image; (5) Sub-step S2-5: based on the camera calibration result, the pixel coordinates of the slider position are converted into physical coordinates in the world coordinate system, so as to subsequently perform motion trajectory reconstruction and ice friction coefficient calculation. (c) Step S3: calculating the motion acceleration of the slider, characterized in that step S3 comprises the following sub-steps: (1) Sub-step S3-1 - According to the kinematic model of the slider, the acceleration of the slider motion is pre-set using the Taylor basis function expansion method, that is, the acceleration equation is reconfigured as a(t) = B·R, where x is displacement, t is time, B 1×(n+1) is an n+1-dimensional row vector, where each element b(j) = t j-1 (j = 1, 2,..., n+1) represents a basis function term, R (n+1)×1 = (r1, r2,..., r n+1 ) T is a coefficient term corresponding to the basis function, and n is the order of the basis function expansion; (2) Sub-step S3-2: based on the recognized slider coordinate information, a Tikhonov regularization method is used to solve the coefficient term corresponding to the base function, so as to obtain the slider acceleration equation; (d) Step S4: inversely obtaining the ice friction coefficient based on the relationship between acceleration and friction mechanics model, characterized in that step S4 comprises the following sub-steps: (1) Sub-step S4-1: assuming that the Coulomb friction model is satisfied between the slider and the ice surface, the mechanical relationship can be expressed as f = μ·N, wherein μ is the ice friction coefficient, and N is the support force of the ice surface on the slider; (2) Sub-step S4-2 - Establish the relationship between the ice friction coefficient and the acceleration according to Newton's second law, i.e. μ = a·g -1 where g is the acceleration due to gravity, thereby achieving non-contact inversion of the ice friction coefficient.

6. A method of measuring ice friction coefficient non-contact as claimed in claim 5, wherein, In the step S1-1, the position distribution of the calibration plate in the image region should include the picture edge, the center, and cover the typical trajectory region of the slider movement.

7. The method of claim 5, wherein the ice friction coefficient is measured without contact. The steps S2 and S3 are realized by a computer program, and all the sub-steps in the respective steps are executed in sequence.

8. The method of claim 5, wherein the ice friction coefficient is measured without contact. In the step S4, the acceleration a calculated in the step S3 is obtained through mathematical operation, so as to complete the non-contact measurement of the ice friction coefficient μ.

Images