Pendulum block motion curve fitting method, device, equipment and readable storage medium

Abstract

The present invention discloses a method, device, equipment and readable storage medium for fitting a pendulum motion curve. The method comprises: acquiring a pendulum motion image, performing feature point detection on a marker in the pendulum motion image to obtain a feature point descriptor set, and performing pixel matching based on the feature point descriptor set to determine a matching pixel point set; calculating the rotational motion center coordinates and rotation angle of the pendulum according to the matching pixel point set, and fitting the motion curve of the pendulum according to the rotational motion center coordinates and rotation angle. The present invention improves the accuracy of the pendulum motion curve fitting by determining the feature point descriptor set of the marker in the pendulum motion image, determining the matching pixel point set based on the feature point descriptor set, and calculating the rotational motion center coordinates and rotation angle of the pendulum according to the matching pixel point set, thereby fitting the motion curve of the pendulum.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and in particular to a pendulum block motion curve fitting method, device, equipment and a readable storage medium. Background Art

[0002] In the study of the dynamic characteristics of centrifugal pendulum dampers, especially when establishing mathematical models for forward analysis, it is necessary to obtain the actual motion of the pendulum block to compensate for the poor fit of the model to the centrifugal pendulum damper's motion. The usual method is to install an angle sensor to measure the motion of the pendulum block and generate its motion curve. On the one hand, the installation of the sensor will have a certain impact on the pendulum block's motion process, causing the generated motion curve to be inconsistent with the actual motion process. On the other hand, the motion of the pendulum block is not a single rotational motion around the central axis of the centrifugal pendulum damper's rotor; it also has a certain amount of self-rotation. Therefore, it is difficult to obtain a relatively accurate motion curve using only an angle sensor. Therefore, how to improve the accuracy of the pendulum block's motion curve fitting is an urgent problem that needs to be solved. Summary of the Invention

[0003] The main purpose of the present invention is to propose a pendulum motion curve fitting method, device, equipment and readable storage medium, aiming to solve the problem of how to improve the accuracy of pendulum motion curve fitting.

[0004] To achieve the above object, the present invention provides a pendulum motion curve fitting method, which comprises the following steps:

[0005] Acquire a motion image of the swing block, perform feature point detection on a marker in the motion image of the swing block to obtain a feature point descriptor set, and perform pixel matching based on the feature point descriptor set to determine a matching pixel point set;

[0006] The rotational motion center coordinates and the rotation angle of the pendulum block are calculated based on the set of matching pixel points, and the motion curve of the pendulum block is fitted based on the rotational motion center coordinates and the rotation angle.

[0007] Preferably, the step of performing feature point detection on the markers in the swing block motion image to obtain a feature point descriptor set includes:

[0008] Acquiring the number of frames of the swing block motion image, and obtaining a swing block motion image set according to the number of frames and the swing block motion image;

[0009] Acquiring position information of a marker in each swing block motion image in the swing block motion image set, and determining a recognition area based on the position information of the marker;

[0010] Two adjacent frames of the swing block motion image are obtained from the swing block motion image set, and feature point detection is performed on the marker in the identification area of the two adjacent frames of the swing block motion image to obtain a feature point descriptor set.

[0011] Preferably, performing pixel matching according to the feature point descriptor set to determine a matching pixel point set includes:

[0012] comparing the feature point descriptors representing two adjacent frames of the swing block motion image in the feature point descriptor set to determine the similarity of the feature point descriptors;

[0013] Pixel matching is performed according to the similarity to determine a matching pixel point set corresponding to the feature point descriptor set in the two adjacent frames of the swing block motion image.

[0014] Preferably, the step of calculating the rotational motion center coordinates and the rotation angle of the pendulum block according to the set of matching pixel points includes:

[0015] Determining homogeneous coordinates according to the set of matching pixel points, and constructing a target transformation matrix according to the set of matching pixel points and the homogeneous coordinates;

[0016] The rotational motion center coordinates and the rotation angle of the pendulum block are calculated based on the matching pixel point set and the target transformation matrix.

[0017] Preferably, the step of constructing a target transformation matrix according to the set of matching pixel points and the homogeneous coordinates includes:

[0018] Determining a transformation process of two adjacent frames of swing block motion images according to the set of matching pixel points, and constructing a transformation matrix set corresponding to the transformation process according to the transformation process and the homogeneous coordinates;

[0019] According to the transformation process, each transformation matrix in the transformation matrix set is sequentially multiplied by the left to obtain a target transformation matrix.

[0020] Preferably, the step of calculating the rotational motion center coordinates and the rotation angle of the pendulum block according to the matching pixel point set and the target transformation matrix includes:

[0021] The matching pixel point matrix set of two adjacent frames of the swing block motion image is determined based on the matching pixel point set, and the rotational motion center coordinates and rotation angle of the swing block are calculated based on the matching pixel point matrix set and the target transformation matrix in combination with the Lagrange multiplier method.

[0022] Preferably, the step of fitting the motion curve of the pendulum block according to the rotational motion center coordinates and the rotation angle includes:

[0023] determining motion curves corresponding to two adjacent frames of the swing block motion image according to the rotational motion center coordinates and the rotation angles corresponding to the two adjacent frames of the swing block motion image;

[0024] The motion curves corresponding to the two adjacent frames of the swing block motion image are spliced to obtain the motion curve of the swing block by fitting.

[0025] In addition, to achieve the above-mentioned purpose, the present invention further provides a pendulum motion curve fitting device, the pendulum motion curve fitting device comprising:

[0026] a determination module, configured to acquire a motion image of a swing block, perform feature point detection on a marker in the motion image of the swing block to obtain a feature point descriptor set, and perform pixel matching based on the feature point descriptor set to determine a matching pixel point set;

[0027] A calculation module is used to calculate the rotational motion center coordinates and the rotation angle of the pendulum block according to the set of matching pixel points, and to fit the motion curve of the pendulum block according to the rotational motion center coordinates and the rotation angle.

[0028] Preferably, the determining module is further configured to:

[0029] Acquiring the number of frames of the swing block motion image, and obtaining a swing block motion image set according to the number of frames and the swing block motion image;

[0030] Acquiring position information of a marker in each swing block motion image in the swing block motion image set, and determining a recognition area based on the position information of the marker;

[0031] Two adjacent frames of the swing block motion image are obtained from the swing block motion image set, and feature point detection is performed on the marker in the identification area of the two adjacent frames of the swing block motion image to obtain a feature point descriptor set.

[0032] Preferably, the determining module is further configured to:

[0033] comparing the feature point descriptors representing two adjacent frames of the swing block motion image in the feature point descriptor set to determine the similarity of the feature point descriptors;

[0034] Pixel matching is performed according to the similarity to determine a matching pixel point set corresponding to the feature point descriptor set in the two adjacent frames of the swing block motion image.

[0035] Preferably, the calculation module is further used for:

[0036] Determining homogeneous coordinates according to the set of matching pixel points, and constructing a target transformation matrix according to the set of matching pixel points and the homogeneous coordinates;

[0037] The rotational motion center coordinates and the rotation angle of the pendulum block are calculated based on the matching pixel point set and the target transformation matrix.

[0038] Preferably, the calculation module is further used for:

[0039] Determining a transformation process of two adjacent frames of swing block motion images according to the set of matching pixel points, and constructing a transformation matrix set corresponding to the transformation process according to the transformation process and the homogeneous coordinates;

[0040] According to the transformation process, each transformation matrix in the transformation matrix set is sequentially multiplied by the left to obtain a target transformation matrix.

[0041] Preferably, the calculation module is further used for:

[0042] The matching pixel point matrix set of two adjacent frames of the swing block motion image is determined based on the matching pixel point set, and the rotational motion center coordinates and rotation angle of the swing block are calculated based on the matching pixel point matrix set and the target transformation matrix in combination with the Lagrange multiplier method.

[0043] Preferably, the calculation module is further used for:

[0044] determining motion curves corresponding to two adjacent frames of the swing block motion image according to the rotational motion center coordinates and the rotation angles corresponding to the two adjacent frames of the swing block motion image;

[0045] The motion curves corresponding to the two adjacent frames of the swing block motion image are spliced to obtain the motion curve of the swing block by fitting.

[0046] In addition, to achieve the above-mentioned purpose, the present invention also provides a pendulum motion curve fitting device, which includes: a memory, a processor, and a pendulum motion curve fitting program stored in the memory and runnable on the processor. When the pendulum motion curve fitting program is executed by the processor, the steps of the pendulum motion curve fitting method described above are implemented.

[0047] In addition, to achieve the above-mentioned purpose, the present invention also provides a readable storage medium, which is a computer-readable storage medium. The readable storage medium stores a pendulum motion curve fitting program, and when the pendulum motion curve fitting program is executed by the processor, the steps of the pendulum motion curve fitting method as described above are implemented.

[0048] The pendulum block motion curve fitting method proposed in the present invention obtains a pendulum block motion image, performs feature point detection on the markers in the pendulum block motion image to obtain a feature point descriptor set, and performs pixel matching based on the feature point descriptor set to determine a matching pixel point set; the rotational motion center coordinates and rotation angle of the pendulum block are calculated based on the matching pixel point set, and the motion curve of the pendulum block is fitted based on the rotational motion center coordinates and rotation angle. The present invention improves the accuracy of the pendulum block motion curve fitting by determining the feature point descriptor set of the markers in the pendulum block motion image, determining the matching pixel point set based on the feature point descriptor set, and calculating the rotational motion center coordinates and rotation angle of the pendulum block based on the axis coordinates and the matching pixel point set. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 Schematic diagram of the device structure of the hardware operating environment involved in the embodiment of the present invention;

[0050] Figure 2 Schematic diagram of the flow of the first embodiment of the pendulum block motion curve fitting method of the present invention.

[0051] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and embodiments. DETAILED DESCRIPTION

[0052] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0053] like Figure 1 As shown, Figure 1 It is a schematic diagram of the device structure of the hardware operating environment involved in the embodiment of the present invention.

[0054] The device in the embodiment of the present invention may be a PC or a server device.

[0055] like Figure 1As shown, the device may include: a processor 1001, such as a CPU, a network interface 1004, a user interface 1003, a memory 1005, and a communication bus 1002. Among them, the communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display), an input unit such as a keyboard (Keyboard), and the user interface 1003 may also include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a WI-FI interface). The memory 1005 may be a high-speed RAM memory, or a stable memory (non-volatile memory), such as a disk memory. The memory 1005 may also be a storage device independent of the aforementioned processor 1001.

[0056] Those skilled in the art will understand that Figure 1 The device structure shown in the figure does not constitute a limitation of the device, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0057] like Figure 1 As shown, the memory 1005 as a computer storage medium may include an operating system, a network communication module, a user interface module, and a pendulum motion curve fitting program.

[0058] Among them, the operating system is a program that manages and controls the portable pendulum motion curve fitting device and software resources, and supports the operation of the network communication module, user interface module, pendulum motion curve fitting program and other programs or software; the network communication module is used to manage and control the network interface 1004; the user interface module is used to manage and control the user interface 1003.

[0059] exist Figure 1 In the pendulum motion curve fitting device shown, the pendulum motion curve fitting device calls the pendulum motion curve fitting program stored in the memory 1005 through the processor 1001 and executes the operations in each embodiment of the pendulum motion curve fitting method described below.

[0060] Based on the above hardware structure, an embodiment of the pendulum block motion curve fitting method of the present invention is proposed.

[0061] Reference Figure 2 , Figure 2 1 is a flow chart of a first embodiment of a method for fitting a motion curve of a pendulum block according to the present invention, the method comprising:

[0062] Step S10: acquiring a motion image of the swing block, performing feature point detection on a marker in the motion image of the swing block to obtain a feature point descriptor set, and performing pixel matching based on the feature point descriptor set to determine a matching pixel point set;

[0063] Step S20 , calculating the rotational motion center coordinates and the rotation angle of the pendulum block according to the set of matching pixel points, and fitting the motion curve of the pendulum block according to the rotational motion center coordinates and the rotation angle.

[0064] The pendulum block motion curve fitting method of this embodiment is applied to a pendulum block motion curve fitting device, which may be a terminal or a PC device. The pendulum block motion curve fitting device includes but is not limited to a high-speed camera with a high frame rate, which is used to capture the motion image of the pendulum block in the centrifugal pendulum vibration absorber. For the convenience of description, the pendulum block motion curve fitting device is taken as an example for description; the pendulum block motion curve fitting device obtains the pendulum block motion image in the centrifugal pendulum vibration absorber, obtains the frame number of the pendulum block motion image, and obtains a pendulum block motion image set based on the frame number and the pendulum block motion image; the pendulum block motion curve fitting device obtains the position information of the marker in each pendulum block motion image in the pendulum block motion image set, determines the identification area based on the position information of the marker, and obtains the pendulum block motion images of two adjacent frames in the pendulum block motion image set. , and perform feature point detection on the markers in the identification area of the two adjacent frames of the pendulum motion image to obtain a feature point descriptor set; the pendulum motion curve fitting device compares the feature point descriptors representing the two adjacent frames of the pendulum motion image in the feature point descriptor set to determine the similarity of the feature point descriptors, and performs pixel matching based on the similarity to determine the matching pixel point set corresponding to the feature point descriptor set in the two adjacent frames of the pendulum motion image; the pendulum motion curve fitting device determines the homogeneous coordinates based on the matching pixel point set, and constructs the target transformation matrix based on the matching pixel point set and the homogeneous coordinates, calculates the rotation center coordinates and rotation angle of the pendulum based on the matching pixel point set and the target transformation matrix, and fits the motion curve of the pendulum based on the rotation center coordinates and rotation angle. It should be noted that the centrifugal pendulum vibration absorber (CPVA) can be used to absorb the torsional vibration caused by the engine speed change movement, and can also be used to transmit torque between the engine output end and the transmission input end.

[0065] The pendulum block motion curve fitting method of this embodiment obtains a pendulum block motion image, performs feature point detection on the markers in the pendulum block motion image to obtain a feature point descriptor set, and performs pixel matching based on the feature point descriptor set to determine a matching pixel point set; the rotational motion center coordinates and rotation angle of the pendulum block are calculated based on the matching pixel point set, and the motion curve of the pendulum block is fitted based on the rotational motion center coordinates and rotation angle. The present invention improves the accuracy of the pendulum block motion curve fitting by determining the feature point descriptor set of the markers in the pendulum block motion image, determining the matching pixel point set based on the feature point descriptor set, and calculating the rotational motion center coordinates and rotation angle of the pendulum block based on the axis coordinates and the matching pixel point set.

[0066] The following describes each step in detail:

[0067] Step S10: acquiring a motion image of the swing block, performing feature point detection on a marker in the motion image of the swing block to obtain a feature point descriptor set, and performing pixel matching based on the feature point descriptor set to determine a matching pixel point set;

[0068] In this embodiment, the pendulum motion curve fitting device captures the pendulum motion image in the centrifugal pendulum shock absorber through a high-speed camera with a high frame rate, obtains the pendulum motion image, performs feature point detection on the marker in the pendulum motion image, obtains a feature point descriptor set, and performs pixel matching based on the feature point descriptor set to determine the matching pixel point set. It can be understood that relevant researchers attach markers to the pendulum of the centrifugal pendulum shock absorber before conducting research. The basic requirements for the markers are that the background color and the color of the marker are as different as possible, and the corner points of the marker are as many and as dispersed as possible. Solid five-pointed stars can be selected to facilitate the pendulum motion curve fitting device to perform feature point detection on the marker in the pendulum motion image.

[0069] Specifically, the step of performing feature point detection on the markers in the swing block motion image to obtain a feature point descriptor set includes:

[0070] Step a, obtaining the frame number of the swing block motion image, and obtaining a swing block motion image set according to the frame number and the swing block motion image;

[0071] In this step, after the pendulum motion curve fitting device obtains the pendulum motion image in the centrifugal pendulum shock absorber, it obtains the frame number of the pendulum motion image, splits the pendulum motion image according to the frame number of the pendulum motion image, and obtains a pendulum motion image set. It can be understood that the pendulum motion image is a video of the pendulum motion process, and the video is composed of frames of images, so there is a correlation between every two frames of the pendulum motion image.

[0072] Step b, obtaining position information of a marker in each swing block motion image in the swing block motion image set, and determining a recognition area based on the position information of the marker;

[0073] In this step, the pendulum motion curve fitting device obtains the position information of the marker in each pendulum motion image in the pendulum motion image set, and determines the recognition area based on the marker's position information; for example, the pendulum motion curve fitting device scans each pendulum motion image in the pendulum motion image set, and determines the position information of the marker in each pendulum motion image based on the preset marker, and determines the recognition area based on the preset recognition area size and the marker's position information; by determining the recognition area, the pendulum motion curve fitting device can specifically identify the marker in the recognition area, reducing computing resource consumption. It should be noted that the recognition area is required to completely contain the marker, and the marker is as close to the center of the recognition area as possible.

[0074] Step c: acquiring two adjacent frames of the swing block motion image set, and performing feature point detection on the markers in the identification area of the two adjacent frames of the swing block motion image to obtain a feature point descriptor set.

[0075] In this step, the oscillating block motion curve fitting device obtains two adjacent frames of oscillating block motion images in the oscillating block motion image set, and performs feature point detection on the markers in the identification areas of the two adjacent frames of oscillating block motion images to obtain a feature point descriptor set. In one embodiment, assuming that the oscillating block motion image set contains 800 oscillating block motion images, the oscillating block motion curve fitting device performs feature point detection on the markers in the identification areas of the two adjacent frames of oscillating block motion images in the oscillating block motion image set according to the time sequence of the oscillating block motion images, that is, firstly performs feature point detection on the markers in the identification areas of the oscillating block motion images of the first frame and the second frame to obtain feature point descriptor sets corresponding to the markers of the oscillating block motion images of the first frame and the second frame, then performs feature point detection on the markers in the identification areas of the oscillating block motion images of the second frame and the third frame to obtain feature point descriptor sets corresponding to the markers of the oscillating block motion images of the second frame and the third frame, and so on, until the feature point descriptor sets corresponding to the markers of the oscillating block motion images of the 799th frame and the 800th frame are obtained. It should be noted that, since the marker will rotate with the pendulum, the feature point detection of the marker in the recognition area of two adjacent frames of the pendulum motion image is performed using the ORB descriptor with good rotation invariance.

[0076] Specifically, performing pixel matching according to the feature point descriptor set to determine a matching pixel point set includes:

[0077] Step d, comparing the feature point descriptors representing two adjacent frames of the swing block motion image in the feature point descriptor set to determine the similarity of the feature point descriptors;

[0078] In this step, the pendulum motion curve fitting device compares the feature point descriptors representing two adjacent frames of pendulum motion images in the feature point descriptor set to determine the similarity of the feature point descriptors. For example, the pendulum motion curve fitting device compares the feature point descriptors corresponding to the first frame of the pendulum motion image with the feature point descriptors corresponding to the second frame of the pendulum motion image in the feature descriptor set. For example, if the marker is a five-pointed star, each pendulum motion image corresponds to five feature point descriptors. The pendulum motion curve fitting device calculates the HAMMING distance (Hamming distance) between the first feature point descriptor of the first frame of the pendulum motion image and the five feature point descriptors of the second frame of the pendulum motion image, determines the similarity based on the HAMMING distance, and then calculates the HAMMING distance between the second feature point descriptor of the first frame of the pendulum motion image and the five feature point descriptors of the second frame of the pendulum motion image, determines the similarity based on the HAMMING distance, and so on, to obtain the similarity between each feature point descriptor of the first frame of the pendulum motion image and each feature point descriptor of the second frame of the pendulum motion image.

[0079] Step e: performing pixel matching according to the similarity to determine a matching pixel set corresponding to the feature point descriptor set in the two adjacent frames of the swing block motion image.

[0080] In this step, the pendulum motion curve fitting device performs pixel matching according to the degree of similarity to determine the set of matching pixels corresponding to the feature point descriptor set in the two adjacent frames of the pendulum motion image; for example: after the pendulum motion curve fitting device determines the degree of similarity of the feature point descriptors corresponding to the first frame and the second frame of the pendulum motion image, it performs pixel matching according to the degree of similarity of the feature point descriptors through the BF matching algorithm (brute force matching algorithm), that is, the pixel corresponding to the feature point descriptor of the first frame of the pendulum motion image and the pixel corresponding to the feature point descriptor with the highest degree of similarity in the second frame of the pendulum motion image are the same pixel, thereby obtaining the coordinate values of the pixel corresponding to each feature point descriptor on the first frame and the second frame of the pendulum motion image, that is, obtaining the set of matching pixels.

[0081] Step S20 , calculating the rotational motion center coordinates and the rotation angle of the pendulum block according to the set of matching pixel points, and fitting the motion curve of the pendulum block according to the rotational motion center coordinates and the rotation angle.

[0082] In this embodiment, the pendulum block motion curve fitting device calculates the rotational motion center coordinates and rotation angle of the pendulum block in two adjacent frames of pendulum block motion images based on the matching pixel point sets of the pendulum block motion images, and determines the motion curves corresponding to the two adjacent frames of pendulum block motion images based on the rotational motion center coordinates and rotation angles corresponding to the two adjacent frames of pendulum block motion images, and then splices the motion curves corresponding to all two adjacent frames of pendulum block motion images to fit the motion curve of the pendulum block.

[0083] Specifically, the step of calculating the rotational motion center coordinates and the rotation angle of the pendulum block according to the set of matching pixel points includes:

[0084] Step f, determining homogeneous coordinates according to the set of matching pixel points, and constructing a target transformation matrix according to the set of matching pixel points and the homogeneous coordinates;

[0085] In this step, the pendulum motion curve fitting device determines the first homogeneous coordinates corresponding to the matching pixel set of the first frame of the pendulum motion image based on the matching pixel set of the first frame of the pendulum motion image in the two adjacent frames of the pendulum motion image, and determines the second homogeneous coordinates corresponding to the matching pixel set of the second frame of the pendulum motion image based on the matching pixel set of the second frame of the pendulum motion image in the two adjacent frames of the pendulum motion image, and constructs the target transformation matrix based on the matching pixel set, the first homogeneous coordinates and the second homogeneous coordinates. It should be noted that homogeneous coordinates are a very useful basic concept in computer graphics. By adding an extra dimension, they can be used to perform matrix transformations on geometric bodies such as scaling, rotation, translation, and perspective projection.

[0086] Furthermore, step f includes:

[0087] Step f1, determining the transformation process of two adjacent frames of the swing block motion image based on the set of matching pixel points, and constructing a transformation matrix set corresponding to the transformation process based on the transformation process and the homogeneous coordinates;

[0088] In this step, the pendulum motion curve fitting device sets the rotation axis projection coordinates of the pendulum of the centrifugal pendulum shock absorber, and determines the transformation process of two adjacent frames of the pendulum motion image based on the matching pixel point set and the rotation axis projection coordinates. Based on the transformation process and the homogeneous coordinates corresponding to each frame of the pendulum motion image in the two adjacent frames of the pendulum motion image, a transformation matrix set corresponding to the transformation process of the two adjacent frames of the pendulum motion image is constructed. For example: let the rotation axis projection coordinates of the pendulum of the centrifugal pendulum shock absorber be point (a, b) in the image, and let X be all matching pixel points (x, y) corresponding to the first frame of the pendulum motion image. T X' is the set of all matching pixels (x', y') in the motion area of the second frame image. TThe matching pixel point refers to the pixel point of the same point on the same object in two adjacent frames of the swing block motion image. The homogeneous coordinate of X is (x, y, 1) T , the homogeneous coordinates of X' are (x', y', 1) T The transformation process of two adjacent frames of swing block motion images can be divided into the following three basic transformations. The first step is to transform the first frame of the swing block motion image to (-a, -b) T Translation; the second step is to rotate the motion image of the pendulum block with the coordinate origin as the rotation center by an angle of θ (clockwise is the positive direction); the third step is to rotate the motion image of the pendulum block to (a, b) T The second frame of the swing block motion image is obtained by translation. The transformation matrix in homogeneous coordinates corresponding to the three-step basic transformation is:

[0089]

[0090]

[0091]

[0092] According to the transformation matrix in homogeneous coordinates corresponding to the three-step basic transformation, a set of transformation matrices corresponding to the transformation process is obtained.

[0093] Step f2: According to the transformation process, each transformation matrix in the transformation matrix set is sequentially multiplied by the left to obtain a target transformation matrix.

[0094] In this step, the pendulum motion curve fitting device multiplies each transformation matrix in the transformation matrix set in the order of the transformation book according to the transformation process of the two adjacent frames of the pendulum motion image to obtain the target transformation matrix; for example, the transformation process of the two adjacent frames of the pendulum motion image can be divided into the following three basic transformations. The first step is to transform the first frame of the pendulum motion image to (-a, -b) T Translation; the second step is to rotate the motion image of the pendulum block with the coordinate origin as the rotation center by an angle of θ (clockwise is the positive direction); the third step is to rotate the motion image of the pendulum block to (a, b) T The second frame of the pendulum motion image is obtained by translation. The pendulum motion curve fitting device multiplies the transformation matrix corresponding to the first step by the transformation matrix corresponding to the second step and then by the transformation matrix corresponding to the third step according to the transformation order of the three-step basic transformation. Finally, the target transformation matrix corresponding to the transformation process of the two adjacent frames of the pendulum motion image is obtained:

[0095]

[0096] Step g: Calculate the rotational motion center coordinates and rotation angle of the pendulum block based on the matching pixel point set and the target transformation matrix.

[0097] In this step, the pendulum motion curve fitting device constructs the first equation based on the matching pixel set of two adjacent frames of the pendulum motion image and the target transformation matrix. Then, the Lagrangian method is used to introduce auxiliary equations to optimize and solve the first equation, and the coordinates of the rotational center and rotation angle of the pendulum are calculated.

[0098] Specifically, step g includes:

[0099] Step g1, determining the matching pixel point matrix set of two adjacent frames of the swing block motion image based on the matching pixel point set, and calculating the rotational motion center coordinates and rotation angle of the swing block based on the matching pixel point matrix set and the target transformation matrix in combination with the Lagrange multiplier method.

[0100] In this step, the pendulum motion curve fitting device determines the matching pixel matrix set of two adjacent frames of the pendulum motion image based on the matching pixel point set, and calculates the rotation center coordinates and rotation angle of the pendulum based on the matching pixel matrix set and the target transformation matrix in combination with the Lagrange multiplier method. For example, the pendulum motion curve fitting device assumes the target transformation matrix E:

[0101]

[0102] in:

[0103] e1=cosθ, e2=sinθ, e3=-acosθ-bsinθ+a, e4=asinθ-bcosθ+b

[0104] Then the transformation process of two adjacent frames of swing block motion images is the corresponding first equation:

[0105] X′=EX

[0106] Determine the matching pixel matrix sets X and X′ of two adjacent frames of the swing block motion image and the i-th pair of matching points x i , x' i The expressions are as follows:

[0107]

[0108]

[0109] Through calculation, we can get the linear equation about vector e, where e is

[0110] e=[e1 e2 e3 e4] T

[0111] Then the first equation can be transformed into:

[0112]

[0113] The first equation is the optimal solution for e. Since the solution of e is a least squares problem with equality constraints; the equality constraints come from the square sum constraints of the cosine and sine of e1 and e2, the least squares problem can be expressed as:

[0114] ste T Λe=1,where

[0115] Use Lagrange multiplication to optimize and solve the above problem, and introduce auxiliary equations:

[0116]

[0117] Derivative the auxiliary equation:

[0118]

[0119] After taking the derivative of the auxiliary equation and simplifying it, we can get:

[0120]

[0121] From the above formula, we can see that for any λ, the solution values of e3 and e4 remain unchanged; therefore, we can set λ = 0 and solve for e3 and e4.

[0122] After taking the derivative of the auxiliary equation and simplifying it, we can get:

[0123]

[0124] After sorting:

[0125]

[0126] set up:

[0127]

[0128]

[0129] Among them, a ij with b i are all known values, so we can get:

[0130]

[0131] Simplified:

[0132]

[0133] The above equations are trigonometric function equations about θ, from which we can solve e1 and e2; since e1, e2, e3 and e4 have been determined, we can combine e1, e2, e3 and e4 into the following formula:

[0134] e1=cosθ, e2=sinθ, e3=-acosθ-bsinθ+a, e4=asinθ-bcosθ+b

[0135] The coordinates of the rotational motion center (a, b) and the rotation angle θ of the pendulum block can be calculated.

[0136] Specifically, the step of fitting the motion curve of the pendulum block according to the rotational motion center coordinates and the rotation angle includes:

[0137] Step h, determining the motion curves corresponding to the two adjacent frames of the swing block motion image according to the rotational motion center coordinates and the rotation angles corresponding to the two adjacent frames of the swing block motion image;

[0138] In this step, the pendulum motion curve fitting device determines the motion curves corresponding to two adjacent frames of pendulum motion images based on the rotational motion center coordinates and rotation angles corresponding to the two adjacent frames of pendulum motion images in the pendulum motion image; for example: assuming that the pendulum motion image set contains 800 pendulum motion images, the pendulum motion curve fitting device first calculates the rotational motion center coordinates and rotation angles corresponding to the first and second frames of the pendulum motion images, and then determines the motion curves corresponding to the first and second frames of the pendulum motion images, and then calculates the rotational motion center coordinates and rotation angles corresponding to the second and third frames of the pendulum motion images, and then determines the motion curves corresponding to the second and third frames of the pendulum motion images, and so on, until the motion curves corresponding to the 799th and 800th frames of the pendulum motion images are determined.

[0139] Step i: splicing the motion curves corresponding to all two adjacent frames of the swing block motion image to obtain the motion curve of the swing block by fitting.

[0140] In this step, the pendulum block motion curve fitting device splices the motion curves corresponding to all two adjacent frames of the pendulum block motion images to fit the motion curve of the pendulum block. For example, assuming that the pendulum block motion image set contains 800 pendulum block motion images, the pendulum block motion curve fitting device splices the motion curves corresponding to the first and second frames of the pendulum block motion images with the motion curves corresponding to the second and third frames of the pendulum block motion images, until the motion curves corresponding to the 799th and 800th frames of the pendulum block motion images are spliced, and finally the motion curve of the pendulum block is fitted.

[0141] The pendulum motion curve fitting device of this embodiment obtains the pendulum motion image in the centrifugal pendulum shock absorber, obtains the frame number of the pendulum motion image, and obtains a pendulum motion image set according to the frame number and the pendulum motion image; the pendulum motion curve fitting device obtains the position information of the marker in each pendulum motion image in the pendulum motion image set, determines the recognition area based on the position information of the marker, obtains two adjacent frames of pendulum motion images in the pendulum motion image set, and performs feature point detection on the markers in the recognition area in the two adjacent frames of the pendulum motion images to obtain a feature point descriptor set; the pendulum motion curve fitting device respectively represents the markers in the feature point descriptor set. The feature point descriptors of two adjacent frames of the pendulum block motion image are compared to determine the similarity of the feature point descriptors, and pixel matching is performed based on the similarity to determine the matching pixel point set corresponding to the feature point descriptor set in the two adjacent frames of the pendulum block motion image; the pendulum block motion curve fitting device determines the homogeneous coordinates based on the matching pixel point set, and constructs the target transformation matrix based on the matching pixel point set and the homogeneous coordinates, and calculates the rotation center coordinates and rotation angle of the pendulum block based on the matching pixel point set and the target transformation matrix, and fits the motion curve of the pendulum block based on the rotation center coordinates and the rotation angle, thereby improving the accuracy of the pendulum block motion curve fitting.

[0142] The present invention also provides a pendulum motion curve fitting device. The pendulum motion curve fitting device of the present invention comprises:

[0143] a determination module, configured to acquire a motion image of a swing block, perform feature point detection on a marker in the motion image of the swing block to obtain a feature point descriptor set, and perform pixel matching based on the feature point descriptor set to determine a matching pixel point set;

[0144] A calculation module is used to calculate the rotational motion center coordinates and the rotation angle of the pendulum block according to the set of matching pixel points, and to fit the motion curve of the pendulum block according to the rotational motion center coordinates and the rotation angle.

[0145] Preferably, the determining module is further configured to:

[0146] Acquiring the number of frames of the swing block motion image, and obtaining a swing block motion image set according to the number of frames and the swing block motion image;

[0147] Acquiring position information of a marker in each swing block motion image in the swing block motion image set, and determining a recognition area based on the position information of the marker;

[0148] Two adjacent frames of the swing block motion image are obtained from the swing block motion image set, and feature point detection is performed on the marker in the identification area of the two adjacent frames of the swing block motion image to obtain a feature point descriptor set.

[0149] Preferably, the determining module is further configured to:

[0150] comparing the feature point descriptors representing two adjacent frames of the swing block motion image in the feature point descriptor set to determine the similarity of the feature point descriptors;

[0151] Pixel matching is performed according to the similarity to determine a matching pixel point set corresponding to the feature point descriptor set in the two adjacent frames of the swing block motion image.

[0152] Preferably, the calculation module is further used for:

[0153] Determining homogeneous coordinates according to the set of matching pixel points, and constructing a target transformation matrix according to the set of matching pixel points and the homogeneous coordinates;

[0154] The rotational motion center coordinates and the rotation angle of the pendulum block are calculated based on the matching pixel point set and the target transformation matrix.

[0155] Preferably, the calculation module is further used for:

[0156] Determining a transformation process of two adjacent frames of swing block motion images according to the set of matching pixel points, and constructing a transformation matrix set corresponding to the transformation process according to the transformation process and the homogeneous coordinates;

[0157] According to the transformation process, each transformation matrix in the transformation matrix set is sequentially multiplied by the left to obtain a target transformation matrix.

[0158] Preferably, the calculation module is further used for:

[0159] The matching pixel point matrix set of two adjacent frames of the swing block motion image is determined based on the matching pixel point set, and the rotational motion center coordinates and rotation angle of the swing block are calculated based on the matching pixel point matrix set and the target transformation matrix in combination with the Lagrange multiplier method.

[0160] Preferably, the calculation module is further used for:

[0161] determining motion curves corresponding to two adjacent frames of the swing block motion image according to the rotational motion center coordinates and the rotation angles corresponding to the two adjacent frames of the swing block motion image;

[0162] The motion curves corresponding to the two adjacent frames of the swing block motion image are spliced to obtain the motion curve of the swing block by fitting.

[0163] The present invention also provides a pendulum block motion curve fitting device.

[0164] The pendulum motion curve fitting device of the present invention includes: a memory, a processor, and a pendulum motion curve fitting program stored in the memory and runnable on the processor. When the pendulum motion curve fitting program is executed by the processor, the steps of the pendulum motion curve fitting method described above are implemented.

[0165] The method implemented when the pendulum motion curve fitting program running on the processor is executed can refer to the various embodiments of the pendulum motion curve fitting method of the present invention, and will not be described in detail here.

[0166] The present invention also provides a readable storage medium.

[0167] The readable storage medium of the present invention stores a pendulum motion curve fitting program, and when the pendulum motion curve fitting program is executed by a processor, the steps of the pendulum motion curve fitting method described above are implemented.

[0168] The method implemented when the pendulum motion curve fitting program running on the processor is executed can refer to the various embodiments of the pendulum motion curve fitting method of the present invention, and will not be described in detail here.

[0169] It should be noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or system comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or system. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or system comprising the element.

[0170] The serial numbers of the above embodiments of the present invention are for description only and do not represent the advantages or disadvantages of the embodiments.

[0171] Through the description of the above embodiments, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus the necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better embodiment. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes a number of instructions for enabling a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods described in each embodiment of the present invention.

[0172] The above are only preferred embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention description and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for fitting a pendulum motion curve, characterized in that: The pendulum block motion curve fitting method comprises the following steps: Acquire a motion image of the swing block, perform feature point detection on a marker in the motion image of the swing block to obtain a feature point descriptor set, and perform pixel matching based on the feature point descriptor set to determine a matching pixel point set; Calculating the rotational motion center coordinates and the rotation angle of the pendulum block according to the set of matching pixel points, and fitting the motion curve of the pendulum block according to the rotational motion center coordinates and the rotation angle; The step of performing pixel matching according to the feature point descriptor set to determine a matching pixel point set includes: comparing the feature point descriptors representing two adjacent frames of the swing block motion image in the feature point descriptor set, and determining the similarity of the feature point descriptors by calculating the Hamming distance between the feature point descriptors of the two adjacent frames of the swing block motion image; Perform pixel matching based on the similarity of the feature point descriptors using a brute force matching algorithm to determine a set of matching pixels corresponding to the feature point descriptor set in two adjacent frames of the swing block motion image; The step of calculating the rotational motion center coordinates and the rotation angle of the pendulum block according to the set of matching pixel points includes: Determining homogeneous coordinates according to the set of matching pixel points, and constructing a target transformation matrix according to the set of matching pixel points and the homogeneous coordinates; The rotational motion center coordinates and the rotation angle of the pendulum block are calculated based on the matching pixel point set and the target transformation matrix.

2. The pendulum motion curve fitting method according to claim 1, wherein: The step of detecting feature points of the markers in the swing block motion image to obtain a feature point descriptor set includes: Acquiring the number of frames of the swing block motion image, and obtaining a swing block motion image set according to the number of frames and the swing block motion image; Acquiring position information of a marker in each swing block motion image in the swing block motion image set, and determining a recognition area based on the position information of the marker; Two adjacent frames of the swing block motion image are obtained from the swing block motion image set, and feature point detection is performed on the marker in the identification area of the two adjacent frames of the swing block motion image to obtain a feature point descriptor set.

3. The pendulum motion curve fitting method according to claim 1, wherein: The step of constructing a target transformation matrix according to the set of matching pixel points and the homogeneous coordinates includes: Determining a transformation process of two adjacent frames of swing block motion images according to the set of matching pixel points, and constructing a transformation matrix set corresponding to the transformation process according to the transformation process and the homogeneous coordinates; According to the transformation process, each transformation matrix in the transformation matrix set is sequentially multiplied by the left to obtain a target transformation matrix.

4. The pendulum motion curve fitting method according to claim 1, wherein: The step of calculating the rotational motion center coordinates and the rotation angle of the pendulum block according to the matching pixel point set and the target transformation matrix includes: The matching pixel point matrix set of two adjacent frames of the swing block motion image is determined based on the matching pixel point set, and the rotational motion center coordinates and rotation angle of the swing block are calculated based on the matching pixel point matrix set and the target transformation matrix in combination with the Lagrange multiplier method.

5. The pendulum motion curve fitting method according to claim 1, wherein: The step of fitting the motion curve of the pendulum block according to the rotational motion center coordinates and the rotation angle comprises: determining motion curves corresponding to two adjacent frames of the swing block motion image according to the rotational motion center coordinates and the rotation angles corresponding to the two adjacent frames of the swing block motion image; The motion curves corresponding to the two adjacent frames of the swing block motion image are spliced to obtain the motion curve of the swing block by fitting.

6. A pendulum motion curve fitting device, characterized in that: The pendulum block motion curve fitting device comprises: A determination module is configured to acquire a swing block motion image, perform feature point detection on a marker in the swing block motion image to obtain a feature point descriptor set, and perform pixel matching based on the feature point descriptor set to determine a matching pixel point set; the determination module is further configured to perform pixel matching based on the feature point descriptor set to determine a matching pixel point set, comprising: comparing feature point descriptors representing two adjacent frames of the swing block motion image in the feature point descriptor set, calculating the Hamming distance between the feature point descriptors of the two adjacent frames of the swing block motion image, to determine the similarity of the feature point descriptors; and performing pixel matching based on the similarity of the feature point descriptors using a brute force matching algorithm to determine a matching pixel point set corresponding to the feature point descriptor set in the two adjacent frames of the swing block motion image; A calculation module is used to calculate the rotational motion center coordinates and rotation angle of the pendulum block based on the matching pixel point set, and fit the motion curve of the pendulum block based on the rotational motion center coordinates and the rotation angle; the calculation module is also used for the steps of calculating the rotational motion center coordinates and rotation angle of the pendulum block based on the matching pixel point set, including: determining homogeneous coordinates based on the matching pixel point set, and constructing a target transformation matrix based on the matching pixel point set and the homogeneous coordinates; calculating the rotational motion center coordinates and rotation angle of the pendulum block based on the matching pixel point set and the target transformation matrix.

7. A pendulum motion curve fitting device, characterized in that: The pendulum motion curve fitting device includes: a memory, a processor, and a pendulum motion curve fitting program stored in the memory and executable on the processor. When the pendulum motion curve fitting program is executed by the processor, the steps of the pendulum motion curve fitting method according to any one of claims 1 to 5 are implemented.

8. A readable storage medium, characterized in that: The readable storage medium is a computer-readable storage medium, and a pendulum motion curve fitting program is stored on the readable storage medium. When the pendulum motion curve fitting program is executed by a processor, the steps of the pendulum motion curve fitting method according to any one of claims 1 to 5 are implemented.

Images