Propeller vibration measurement method based on monocular vision

Abstract

The present invention discloses a propeller vibration measurement method based on monocular vision, which includes: attaching a cooperative target to the propeller blade surface; using a fill light device to fill light on the measurement scene where the propeller is located; using a fixed high-speed camera to collect images of the high-speed rotating propeller according to a set sampling frame rate; processing the collected images, extracting the two-dimensional coordinates of the position points of the cooperative target in the image, and obtaining multiple measurement points; based on the obtained multiple measurement points, using an offline vibration detection algorithm to determine whether the propeller has abnormal frequency vibration and the consistency of the propeller rotation plane. The complexity of propeller vibration measurement and the test cost are reduced.

Description

Technical Field

[0001] The invention belongs to the technical field of propeller vibration measurement, and more specifically, relates to a propeller vibration measurement method based on monocular vision. Background Art

[0002] At present, visual measurement technology is based on machine vision and uses images as a means of detecting information. It has the characteristics of high efficiency, high precision, non-contact and long-distance measurement. It is an important supplement to traditional measurement methods and is widely used in industry, military, medicine and other fields.

[0003] Vibration measurement is one of the key technologies in the development of modern manufacturing industry and has important application significance in military, industrial, agricultural and other fields.

[0004] In propeller vibration detection, the impact on aerodynamics is very high, and a non-contact measurement method is required, and machine vision measurement can meet the requirements. In order to achieve non-contact measurement with a large field of view and high efficiency, the existing vibration measurement system based on binocular vision principle needs to stick multiple marker points on the surface of the object to be measured, and perform multi-point synchronous and accurate matching calculations on the marker points on the left and right camera images to obtain their three-dimensional spatial coordinates and realize three-dimensional vibration measurement.

[0005] However, the use of binocular vision technology to measure propeller vibration has the disadvantages of complex methods and high equipment costs. Summary of the invention

[0006] The purpose of the invention is to propose a propeller vibration measurement method based on monocular vision to reduce the complexity and test cost of propeller vibration measurement.

[0007] To achieve the above object, the present invention proposes a propeller vibration measurement method based on monocular vision, comprising:

[0008] attaching the cooperative target to the surface of the propeller blade;

[0009] Use a fill-in light device to fill in the light of the measurement scene where the propeller is located;

[0010] The image of the high-speed rotating propeller is collected by a fixed high-speed camera according to a set sampling frame rate;

[0011] Processing the collected image, extracting the two-dimensional coordinates of the position points of the cooperative target in the image, and obtaining a plurality of measurement points;

[0012] Based on the obtained multiple measurement points, an offline vibration detection algorithm is used to determine whether the propeller has vibrations of abnormal frequency and the consistency of the propeller rotation plane.

[0013] Optionally, the offline vibration detection algorithm includes:

[0014] Calculate the rotation matrix and translation matrix of the camera relative to the cooperative target;

[0015] Calibrate the stationary coordinate system to obtain the propeller absolute coordinate system;

[0016] Converting the measurement points into the propeller coordinate system according to the rotation matrix and the translation matrix;

[0017] By using the fluctuation of Z data in the three-dimensional coordinates of each measuring point in the propeller coordinate system, it is possible to determine whether there is abnormal frequency vibration of the propeller during its rotation and the consistency of the propeller rotation plane.

[0018] Optionally, the step of calculating a rotation matrix and a translation matrix of the camera relative to the cooperative target comprises:

[0019] According to the coordinates of the measurement points, the size of the cooperative target and the camera internal parameters, an iterative PNP algorithm is used to obtain the rotation matrix and the translation matrix of the camera relative to the cooperative target.

[0020] Optionally, calibrating the stationary coordinate system to obtain the propeller coordinate system includes:

[0021] manually rotating the propeller;

[0022] Manually rotate the propeller, and when the propeller stops, collect images of the propeller using the high-speed camera, and repeat this step at least 3 times;

[0023] Acquire the coordinates of the cooperative target in the acquired image to form at least three measurement points;

[0024] Calculate the translation matrix and rotation matrix of the camera relative to the cooperative target at this time according to the obtained measurement points;

[0025] The multiple measurement points are converted into a propeller coordinate system with the propeller as the center and the propeller rotation plane as the xoy plane to obtain the propeller coordinate system.

[0026] Optionally, the offline vibration detection algorithm includes:

[0027] According to the coordinates of the multiple measurement points of the cooperative target in the image, a circular motion trajectory of the cooperative target corresponding to the propeller without vibration is fitted by the least square method, wherein the circular motion trajectory is a standard circle;

[0028] Obtaining vertical distances between different measuring points and the circular trajectory;

[0029] Perform FFT transformation on the acquired distance data, and determine whether there is vibration of abnormal frequency on the propeller rotation plane based on the transformation result.

[0030] Optionally, the cooperative target is an arucomarker target.

[0031] Optionally, the size of the cooperative target is less than or equal to 1 cm 2 , the mapping of the size of the cooperative target to the image size is less than 0.04 cm / pixel.

[0032] Optionally, the sampling frame rate of the high-speed camera is calculated by the following formula:

[0033]

[0034] Among them, Δθ is the resolution of the measuring point in the circle, unit: °, f r is the sampling frame rate, unit: pfs, ω is the rotation speed of the propeller, unit: rpm.

[0035] Optionally, the processing of the acquired image includes:

[0036] Opencv is used to process the acquired images and extract the position coordinates of the target on the image.

[0037] Optionally, while the fill light device is used to fill light the measurement scene where the propeller is located, the method also includes:

[0038] A cooling device is used to cool the fill light device.

[0039] The beneficial effects of the present invention are:

[0040] The present invention sets a cooperative target on the propeller surface as an image acquisition target for monocular vision, uses a fill light device and a high-speed camera to acquire images of the cooperative target, obtains the measurement point coordinates of the cooperative target in the acquired image, and uses an offline vibration detection algorithm based on the measurement point coordinates to calculate whether there is abnormal frequency vibration and the consistency of the rotation plane in the high-speed rotation of the propeller. The vibration of the propeller and the consistency of the rotation plane can be accurately and quickly measured. Compared with the complexity of the existing binocular vision vibration detection technology, the method is simpler, more efficient and has lower cost.

[0041] The system of the present invention has other characteristics and advantages, which will be apparent from the drawings incorporated herein and the following detailed description, or will be described in detail in the drawings incorporated herein and the following detailed description, which together serve to explain the specific principles of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The above and other objects, features and advantages of the present invention will become more apparent through a more detailed description of exemplary embodiments of the present invention in conjunction with the accompanying drawings, in which like reference numerals generally represent like components.

[0043] Figure 1 A step diagram of a propeller vibration measurement method based on monocular vision according to the present invention is shown.

[0044] Figure 2 A schematic diagram of a measuring device used in a propeller vibration measuring method based on monocular vision according to an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0045] The present invention will be described in more detail below with reference to the accompanying drawings. Although preferred embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to make the present invention more thorough and complete, and to fully convey the scope of the present invention to those skilled in the art.

[0046] Figure 1 A step diagram of a propeller vibration measurement method based on monocular vision according to the present invention is shown.

[0047] like Figure 1 As shown, a propeller vibration measurement method based on monocular vision includes:

[0048] Step S1: attaching the cooperative target to the surface of the propeller blade;

[0049] The cooperative target is an ArucoMarker target. Preferably, the size of the cooperative target is less than or equal to 1 cm. 2 , the mapping of the size of the cooperative target to the image size is less than 0.04 cm / pixel.

[0050] In a specific application scenario, since monocular measurement uses known prior information to complete the entire measurement task, corresponding cooperative targets are required. Figure 2 As shown, this method mainly relies on hardware parts such as high-intensity light source, cooperative target, high-speed camera and offline vibration detection algorithm.

[0051] Among them, the cooperative target adopts ArucoMarker target, which can represent actual information in a smaller range. ArucoMarker is a marker similar to a QR code. By inputting the size of the marker, the marker in the image can be detected and the 3D position of the marker in the camera coordinate system can be calculated. In opencv, there is an aruco library, which can be called to realize functions such as marker generation, marker detection, and pose calculation. By estimating the 3D pose of the marker in the camera coordinate system, you only need to know the intrinsic parameters of the camera (including internal parameters and distortion coefficients) and the size of the marker to output the transformation R, t matrix from the marker coordinate system to the camera coordinate system. Visual target detection based on ArucoMarker is a prior art, and the specific principle will not be repeated here.

[0052] In the specific implementation process, the cooperative target needs to be placed on a plane. Since the propeller is generally a streamlined surface, the scale of the target can be reduced, such as scaling to 1 cm or even smaller. In order to ensure the measurement accuracy, it is necessary to ensure that the mapping re (unit: cm / pixel) of the actual size of the target to the image size is as small as possible. 0.04cm / pixel.

[0053] Step S2: using a fill light device to fill light the measurement scene where the propeller is located;

[0054] Preferably, while using the fill-light device to fill-light the measurement scene where the propeller is located, the method also includes: using a cooling device to cool the fill-light device.

[0055] Step S3: using a high-speed camera at a fixed position to collect images of the high-speed rotating propeller according to a set sampling frame rate;

[0056] In the above specific application scenarios, this method needs to improve the image resolution, and needs to select a lens with a suitable focal length according to the actual measurement distance; and use a high-speed camera with a higher resolution. According to the actual engineering selection, it is also possible to choose not to measure the complete propeller rotation cycle, but only to measure part of the propeller rotation cycle to increase the image resolution.

[0057] In the specific implementation process, in order to capture the transient information of the high-speed rotating propeller and reduce the impact of motion on image accuracy, it is necessary to reduce the exposure time of the image as much as possible. While reducing the exposure time, the brightness of the image will gradually decrease, and it may be imaged under natural conditions, so an additional fill light device is required. It is preferred to use a fill light device with a high-intensity light source to fill the scene. The intensity of the light source can be determined according to the exposure time. Generally, due to the large intensity of the required light source, the light source equipment will generate a lot of heat, so a corresponding cooling device is required to ensure the safety of the light source equipment.

[0058] In this step, the sampling frame rate of the high-speed camera is calculated by the following formula:

[0059]

[0060] Among them, Δθ is the resolution of the measuring point in the circle, unit: °, f r is the sampling frame rate, unit: pfs, ω is the rotation speed of the propeller, unit: rpm.

[0061] Specifically, the high-speed camera needs to determine the sampling frame rate f of the high-speed camera according to the rotation speed ω (unit: rpm) of the propeller and the resolution Δθ (unit: °) of the measurement point in the circumference. r (Unit: pfs). The numerical relationship satisfies: For example, when the propeller speed is 2000rpm, if the acquisition frame rate of the high-speed camera is selected to be 2000pfs, the rotation angle of the propeller for image acquisition is 6°. The acquisition frame rate of the high-speed camera needs to be determined according to actual needs.

[0062] Step S4: Processing the collected image, extracting the two-dimensional coordinates of the position points of the cooperative target in the image, and obtaining a plurality of measurement points;

[0063] In the above specific application scenario, opencv is used to process the arucomarker target, extract the coordinates of the target's position on the image, and then the iterative PNP algorithm is performed according to the coordinates of the obtained measurement points to obtain the rotation matrix and translation matrix of the camera relative to the target.

[0064] Step S5: Based on the obtained multiple measurement points, an offline vibration detection algorithm is used to determine whether the propeller has vibrations of abnormal frequency and the consistency of the propeller rotation plane.

[0065] In the above specific application scenario, the offline vibration detection algorithm of step S5 includes:

[0066] Calculate the rotation matrix and translation matrix of the camera relative to the cooperative target;

[0067] Calibrate the stationary coordinate system to obtain the propeller absolute coordinate system;

[0068] Converting the measurement points into the propeller coordinate system according to the rotation matrix and the translation matrix;

[0069] By using the fluctuation of Z data in the three-dimensional coordinates of each measuring point in the propeller coordinate system, it is possible to determine whether there is abnormal frequency vibration of the propeller during its rotation and the consistency of the propeller rotation plane.

[0070] Specifically, the rotation matrix and translation matrix of the camera relative to the propeller obtained by the rotating propeller are converted into the propeller coordinate system. The fluctuation of the Z data in the propeller coordinate system can be used to determine whether there is abnormal frequency vibration during the high-speed rotation of the propeller, as well as the consistency of the rotation plane during the propeller rotation.

[0071] The calculation of the rotation matrix and translation matrix of the camera relative to the cooperative target includes:

[0072] According to the coordinates of the measurement points, the size of the cooperative target and the camera internal parameters, an iterative PNP algorithm is used to obtain the rotation matrix and the translation matrix of the camera relative to the cooperative target.

[0073] Specifically, after extracting the two-dimensional coordinates of the position point of the target on the image, a plurality of two-dimensional measurement points are obtained, and then an iterative PNP algorithm is performed according to the coordinates of the obtained measurement points to obtain the rotation matrix and translation matrix of the camera relative to the target.

[0074] Wherein, the calibrating the stationary coordinate system to obtain the propeller coordinate system includes:

[0075] manually rotating the propeller;

[0076] Manually rotate the propeller, and when it stops (to obtain the accurate position of the propeller coordinate system in a non-interference state), use the high-speed camera to capture an image of the propeller, and repeat this step at least 3 times;

[0077] Acquire the coordinates of the cooperative target in the acquired image to form at least three measurement points;

[0078] Calculate the translation matrix and rotation matrix of the camera relative to the cooperative target at this time according to the obtained measurement points;

[0079] The multiple measurement points are converted into a propeller coordinate system with the propeller as the center and the propeller rotation plane as the xoy plane to obtain the propeller coordinate system.

[0080] Specifically, since the camera and propeller installation positions may not be in the same plane, it is necessary to measure the propeller's vibration-free circular motion coordinate system. Manually rotate the propeller and let it stop naturally, and measure the rotation matrix and translation matrix of the camera relative to the target at this time. At least 3 points can be measured to complete the plane determination. Convert the measurement points into a propeller coordinate system centered on the propeller and with the propeller rotation plane as the xoy plane to obtain the propeller coordinate system.

[0081] In another embodiment, the offline vibration detection algorithm in step S5 may include:

[0082] According to the coordinates of the multiple measurement points of the cooperative target in the image, a circular motion trajectory of the cooperative target corresponding to the propeller without vibration is fitted by the least square method, wherein the circular motion trajectory is a standard circle;

[0083] Obtaining vertical distances between different measuring points and the circular trajectory;

[0084] The acquired distance data is transformed by FFT (Fast Fourier Transform), and based on the transformation result, it is determined whether there is vibration of abnormal frequency on the propeller rotation plane.

[0085] Specifically, if it is not converted to the propeller plane, the following algorithm can also be used to determine whether there is abnormal frequency vibration. If there is no vibration, the motion trajectory of a point of the target should be a standard circle. Therefore, the circular trajectory of the coordinates can be fitted by the least squares method, and by comparing the distance of the actual point perpendicular to the standard circle and performing FFT transformation on the distance, it can be determined whether there is a large vibration on the rotating plane.

[0086] The embodiments of the present invention have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments.

Claims

1. A propeller vibration measurement method based on monocular vision, characterized in that: include: attaching the cooperative target to the surface of the propeller blade; Use a fill-in light device to fill in the light of the measurement scene where the propeller is located; The image of the high-speed rotating propeller is collected by a fixed high-speed camera according to a set sampling frame rate; Processing the collected image, extracting the two-dimensional coordinates of the position points of the cooperative target in the image, and obtaining a plurality of measurement points; Based on the obtained multiple measurement points, an offline vibration detection algorithm is used to determine whether the propeller has vibrations of abnormal frequency and the consistency of the propeller rotation plane; Wherein, the offline vibration detection algorithm includes: Calculate the rotation matrix and translation matrix of the camera relative to the cooperative target; Calibrate the stationary coordinate system to obtain the propeller absolute coordinate system; Converting the measurement points into the propeller coordinate system according to the rotation matrix and the translation matrix; By using the fluctuation of Z data in the three-dimensional coordinates of each measuring point in the propeller coordinate system, it is determined whether there is abnormal frequency vibration of the propeller during the rotation process and the consistency of the propeller rotation plane; The offline vibration detection algorithm also includes: According to the coordinates of the multiple measurement points of the cooperative target in the image, a circular motion trajectory of the cooperative target corresponding to the propeller without vibration is fitted by the least square method, wherein the circular motion trajectory is a standard circle; Obtaining vertical distances between different measuring points and the circular trajectory; Perform FFT transformation on the acquired distance data, and determine whether there is vibration of abnormal frequency on the propeller rotation plane based on the transformation result.

2. The propeller vibration measurement method based on monocular vision according to claim 1, characterized in that: The calculation of the rotation matrix and translation matrix of the camera relative to the cooperative target includes: According to the coordinates of the measurement points, the size of the cooperative target and the camera internal parameters, an iterative PNP algorithm is used to obtain the rotation matrix and the translation matrix of the camera relative to the cooperative target.

3. The propeller vibration measurement method based on monocular vision according to claim 1, characterized in that: The calibrating the stationary coordinate system to obtain the propeller coordinate system includes: Manually rotate the propeller, and when the propeller stops, collect images of the propeller using the high-speed camera, and repeat this step at least 3 times; Acquire the coordinates of the cooperative target in the acquired image to form at least three measurement points; Calculate the translation matrix and rotation matrix of the camera relative to the cooperative target at this time according to the obtained measurement points; The multiple measurement points are converted into a propeller coordinate system with the propeller as the center and the propeller rotation plane as the xoy plane to obtain the propeller coordinate system.

4. The propeller vibration measurement method based on monocular vision according to claim 1, characterized in that: The cooperative target is an arucomarker target.

5. The propeller vibration measurement method based on monocular vision according to claim 4 is characterized in that: The size of the cooperative target is less than or equal to 1 cm 2 , the mapping of the size of the cooperative target to the image size is less than 0.04 cm / pixel.

6. The propeller vibration measurement method based on monocular vision according to claim 1, characterized in that: The sampling frame rate of the high-speed camera is calculated by the following formula: Among them, Δθ is the resolution of the measuring point in the circle, unit: °, f r is the sampling frame rate, unit: pfs, ω is the rotation speed of the propeller, unit: rpm.

7. The propeller vibration measurement method based on monocular vision according to claim 1, characterized in that: The processing of the collected images comprises: Opencv is used to process the acquired images and extract the position coordinates of the target on the image.

8. The propeller vibration measurement method based on monocular vision according to claim 1, characterized in that: While using the fill light device to fill light for the measurement scene where the propeller is located, it also includes: A cooling device is used to cool the fill light device.

Images