Vibrating table test sample real-time deformation monitoring system based on machine vision

Abstract

The invention discloses a vibration table test sample real-time deformation monitoring system based on machine vision. The vibration table test sample real-time deformation monitoring system comprises an active modulation light source module, a multi-camera high-speed acquisition subsystem, a real-time deformation analysis unit, a full-field strain field reconstruction module, a motion compensation fusion module and a vibration synchronous control interface. An active modulation light source is synchronous with vibration, and motion blur is effectively eliminated in cooperation with a narrow pulse width design; the multi-camera orthogonal layout is combined with the vibration isolation base, so that stable acquisition of images is guaranteed; and the high-contrast coding datum point and dynamic tracking module accurately obtains the 6-DOF displacement. The rigid body and the elastic deformation can be accurately separated through a digital image correlation method and motion compensation fusion. High-frequency updating of a full-resolution strain field is achieved through multi-scale feature fusion, and monitoring intuition and safety are improved through an online visualization and early warning module. The system can realize high-precision, real-time and full-field deformation monitoring, meets complex vibration test requirements, and provides reliable data support for material and structure performance evaluation.

Description

Technical Field

[0001] The present invention relates to the technical field of deformation monitoring systems, and particularly to a real-time deformation monitoring system for a vibration table test sample based on machine vision. Background Art

[0002] In vibration table tests, accurately monitoring the real-time deformation of samples is crucial for evaluating the dynamic performance of materials and structures. Current traditional deformation monitoring methods, such as strain gauge measurements and laser displacement sensors, have limitations such as sparse measurement points and difficulty in obtaining full-field strain information; while conventional machine vision monitoring technologies are prone to motion blur interference when facing high-frequency vibrations of samples, resulting in insufficient clarity of image acquisition, and lack of effective discrimination between rigid body motion and elastic deformation, making it impossible to accurately restore the true deformation of the sample. In addition, existing monitoring systems also have obvious shortcomings in terms of real-time data processing and the accuracy of multi-source data fusion, and it is difficult to meet the high-precision and real-time monitoring requirements for sample deformation under complex working conditions. Therefore, there is an urgent need for a system that can effectively overcome motion blur, accurately separate rigid body and elastic deformation, and achieve real-time high-precision monitoring to meet the growing technical needs in the field of vibration table tests. Summary of the Invention

[0003] In order to overcome the deficiencies of the existing technical solutions, the present invention provides a real-time deformation monitoring system for a vibration table test sample based on machine vision, which can effectively solve the problems raised in the background art.

[0004] The technical solution adopted by the present invention to solve its technical problems is as follows:

[0005] A real-time deformation monitoring system for a vibration table test sample based on machine vision, comprising:

[0006] An active modulation light source module: adopting a modifiable light source with high frequency and narrow pulse width, the light emission timing of which is synchronized with the motion phase of the vibration table, and is used to eliminate motion blur during the vibration of the sample;

[0007] A multi-camera high-speed acquisition subsystem: at least two high-frame-rate industrial cameras are fixed on a rigid bracket at orthogonal angles, the bracket is installed independently of the vibration table through a vibration isolation base, each camera is equipped with an optical filter to match the light source wavelength, and is directly connected to the processing unit through a gigabit network port;

[0008] A real-time deformation analysis unit: built-in with a reference point dynamic tracking module, the reference point dynamic tracking module identifies the high-contrast coded reference points pre-placed on the surface of the sample, and real-time calculates its 6-DOF displacement in the vibration coordinate system;

[0009] A full-field strain field reconstruction module: based on the digital image correlation method (DIC), sub-pixel-level displacement vector calculation is performed on the speckles on the surface of the sample;

[0010] Motion Compensation Fusion Module: Utilize the reference point displacement data to perform rigid body motion compensation on the DIC results and output a pure deformation strain field.

[0011] Vibration Synchronization Control Interface: Receive the phase trigger signal from the vibration table controller and dynamically adjust the light source exposure window and camera acquisition timing.

[0012] As a further description of the above technical solution, the active modulation light source module includes an LED array with a pulse width ≤ 1 μs, and the drive circuit of the LED array responds to the synchronization signal sent by the vibration synchronization control interface.

[0013] As a further description of the above technical solution, the reference points are dot matrices made of rare earth fluorescent materials, whose excitation wavelength matches that of the active modulation light source. Each reference point has a unique binary coding pattern, and the spatial resolution ≥ 0.05 mm.

[0014] As a further description of the above technical solution, the real-time deformation analysis unit further includes a multi-scale feature fusion module. The multi-scale feature fusion module performs Kalman filter fusion on the rigid body displacement data tracked by the reference points and the local strain data of DIC in the spatio-temporal domain, outputs a full-resolution strain field, and the update frequency ≥ 500 Hz.

[0015] As a further description of the above technical solution, the motion compensation fusion module performs the following operations:

[0016] (a) Establish a local coordinate system with the reference point group as the reference;

[0017] (b) Map the DIC displacement vector to this local coordinate system through affine transformation;

[0018] (c) Eliminate the rigid body motion component and retain the elastic deformation component.

[0019] As a further description of the above technical solution, the vibration isolation base adopts an air-floating vibration isolation platform. The natural frequency of the air-floating vibration isolation platform ≤ 2 Hz, and the vibration energy transmitted from the vibration table to the camera bracket is attenuated by ≥ 40 dB.

[0020] As a further description of the above technical solution, the vibration synchronization control interface can turn on the light source exposure window and the camera exposure time at the peak moment of the vibration table acceleration, and the deviation between the camera exposure time and the light source pulse width ≤ 0.1 μs.

[0021] As a further description of the above technical solution, it further includes an online visualization terminal and an early warning module. The online visualization terminal displays the deformation heat map, strain evolution curve, and displacement-time history diagram of key positions in real time. The early warning module triggers an alarm signal when the local strain exceeds the material yield threshold.

[0022] As a further description of the above technical solution, the wavelength of the light source is in the near-infrared band located at 700–900 nm, and the passband bandwidth of the optical filter is ≤10 nm.

[0023] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0024] A real-time deformation monitoring system for a vibration table test sample based on machine vision according to the present invention has at least one of the following beneficial effects during use:

[0025] The actively modulated light source is synchronized with the vibration, and in combination with the narrow pulse width design, it effectively eliminates motion blur; the multi-camera orthogonal layout combined with the vibration isolation base ensures stable image acquisition; the high-contrast coded reference points and the dynamic tracking module accurately obtain 6-DOF displacements. Through digital image correlation method and motion compensation fusion, the rigid body and elastic deformations can be accurately separated. The multi-scale feature fusion realizes high-frequency update of the full-resolution strain field, and the online visualization and warning module improves the intuitiveness and safety of monitoring. The system can achieve high-precision, real-time, and full-field deformation monitoring, meet the requirements of complex vibration tests, and provide reliable data support for the evaluation of material and structural properties. Description of the Drawings

[0026] Figure 1 It is a schematic diagram of the overall structure of a real-time deformation monitoring system for a vibration table test sample based on machine vision according to the present invention;

[0027] Figure 2 It is a schematic diagram of the structural framework of a real-time deformation monitoring system for a vibration table test sample based on machine vision according to the present invention;

[0028] Figure 3 It is a processing flow chart of the real-time deformation analysis unit of a real-time deformation monitoring system for a vibration table test sample based on machine vision according to the present invention;

[0029] Figure 4 It is a schematic diagram of the principle of a real-time deformation monitoring system for a vibration table test sample based on machine vision according to the present invention.

[0030] Reference numerals in the drawings:

[0031] 101, actively modulated light source module; 201, multi-camera high-speed acquisition subsystem; 301, rigid support; 401, vibration isolation base. Detailed Embodiments

[0032] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0033] As Figures 1-4 shown, the present invention provides a real-time deformation monitoring system for a vibration table test sample based on machine vision, including:

[0034] Active modulation light source module 101: Adopts a modifiable light source with high frequency and narrow pulse width, and its light emission timing is synchronized with the motion phase of the vibration table, used to eliminate motion blur during the vibration of the sample;

[0035] In this embodiment, under the high-frequency vibration of the vibration table, motion blur is eliminated and the rigid body motion and elastic deformation of the sample are separated to realize real-time monitoring of the full-field strain. An LED array with a pulse width ≤ 1 μs is used to instantaneously expose at the phase point where the motion speed of the vibration table is the smallest / the direction is reversed (such as the acceleration peak). A phase trigger signal (such as a 0° or 180° phase angle) is received through the vibration synchronization control interface to make the exposure window strictly synchronized with the vibration period.

[0036] Multi-camera high-speed acquisition subsystem 201: At least two high-frame-rate industrial cameras are fixed on a rigid bracket 301 at orthogonal angles. The bracket is installed independently of the vibration table through a vibration isolation base 401. Each camera is equipped with an optical filter to match the light source wavelength and is directly connected to the processing unit through a gigabit network port;

[0037] The camera bracket is independent of the vibration table to eliminate measurement errors caused by the jitter of the reference system. The gigabit network direct connection ensures the real-time transmission of 1280×1024 resolution images at a high frame rate (≥500fps).

[0038] Real-time deformation analysis unit: Built-in with a reference point dynamic tracking module, the reference point dynamic tracking module identifies the high-contrast coded reference points pre-placed on the surface of the sample and real-time calculates their 6-DOF displacements in the vibration coordinate system;

[0039] Rare earth fluorescent dots (such as YAG:Ce 3+ ) emit visible light under near-infrared excitation to achieve passive marking under active illumination. A local coordinate system is established through three non-collinear reference points, and translations (X, Y, Z) and rotations (Rx, Ry, Rz) are calculated in real time.

[0040] Full-field strain field reconstruction module: Based on digital image correlation method (DIC), sub-pixel level displacement vector calculation is performed on the speckles on the surface of the sample;

[0041] The sub-pixel DIC algorithm uses the inverse composite Gauss-Newton method (IC-GN) to achieve a displacement accuracy of 0.01 pixel level. The Kalman filter fuses the reference points (low frequency but globally accurate) and DIC data (high frequency but locally noisy) to output a full-resolution strain field. The deviation between the light source exposure and the camera shutter is ≤0.1 μs, and FPGA hardware-level synchronization is required. The exposure window is opened at the peak of the vibration acceleration (when the speed ≈ 0) to maximize the suppression of blurring.

[0042] Motion compensation fusion module: Use the displacement data of the reference points to perform rigid body motion compensation on the DIC results and output a pure deformation strain field;

[0043] Vibration synchronization control interface: Receive the phase trigger signal of the vibration table controller and dynamically adjust the light source exposure window and the camera acquisition timing.

[0044] Traditional continuous light sources produce smear under high-speed vibration (for example, the blurring reaches the millimeter level when vibrating at 50 Hz + 1 ms exposure), while 1 μs exposure can reduce the blurring amount to the micron level. The narrow pulse + narrow-band filter (bandwidth ≤ 10 nm) effectively suppresses ambient light interference, especially suitable for industrial sites. The two cameras are at a 90° angle to each other to eliminate the single-view occlusion problem (such as the blind area when the sample warps). The air-bearing platform (natural frequency ≤ 2 Hz) attenuates vibration energy of ≥ 40 dB (for example, the vibration transmissibility at 100 Hz < 1%), ensuring the absolute stillness of the camera coordinate system. The 700 - 900 nm near-infrared light source avoids visible light interference, and the speckle on the metal surface has a high reflectivity in this wavelength band.

[0045] Furthermore, the active modulation light source module 101 includes an LED array with a pulse width ≤ 1 μs, and the drive circuit of the LED array responds to the synchronization signal sent by the vibration synchronization control interface.

[0046] In this embodiment, μs-level exposure is used to eliminate motion blurring, sub-pixel DIC is used to achieve micro-strain measurement, reference point compensation is used to decouple rigid body motion, and air-bearing vibration isolation is used to ensure the stability of the reference system. Finally, real-time and accurate monitoring of the full-field strain is achieved under 500 Hz high-frequency vibration.

[0047] Furthermore, the reference points are dot matrices made of rare earth fluorescent materials, whose excitation wavelength matches that of the active modulation light source. Each reference point has a unique binary coding pattern, and the spatial resolution ≥ 0.05 mm.

[0048] The LED array receives the vibration synchronization signal through the drive circuit (such as triggered by the 0° or 180° phase angle of the vibration table), and emits narrow pulse light with a pulse width ≤ 1 μs at the peak of the acceleration (when the speed ≈ 0). The exposure window is strictly synchronized with the vibration period to ensure that the instantaneous motion speed of the sample is minimized at the moment of image acquisition.

[0049] Under 100 Hz vibration, traditional 1 ms exposure results in a blur amount > 100 μm, while 1 μs exposure reduces the blur to < 1 μm. Short pulses + synchronous control suppress ambient light, and the image contrast is increased by more than 3 times, providing a reliable input for DIC.

[0050] Furthermore, the real-time deformation analysis unit further includes a multi-scale feature fusion module, which performs Kalman filter fusion on the rigid body displacement data tracked by the reference points and the local strain data of DIC in the spatio-temporal domain, outputs a full-resolution strain field, and the update frequency ≥ 500 Hz.

[0051] Rare earth fluorescent materials (such as YAG:Ce 3+ ) emit visible fluorescence under excitation by 700 - 900 nm near-infrared light, forming high-contrast marking points. The binary coding pattern realizes a unique ID through the spatial arrangement of dot matrices (for example, a 3×3 dot matrix can encode 512 kinds of IDs), and combines with a sub-pixel positioning algorithm (such as the centroid method) to achieve position calculation.

[0052] Furthermore, the motion compensation fusion module performs the following operations:

[0053] (a) Establish a local coordinate system with the reference point group as the reference;

[0054] (b) Map the DIC displacement vector to this local coordinate system through affine transformation;

[0055] (c) Eliminate the rigid body motion component and retain the elastic deformation component.

[0056] The reference points provide low-frequency global displacements (high precision), and DIC provides high-frequency local strains (high resolution). Fusion is achieved by optimizing the covariance matrix through Kalman filtering. A 500 Hz update rate can resolve 250 Hz vibration modes (meeting the Nyquist theorem), and the calculation efficiency is 10 times higher than that of pure DIC.

[0057] Furthermore, the vibration isolation base 401 adopts an air-bearing vibration isolation platform, the natural frequency of the air-bearing vibration isolation platform ≤ 2 Hz, and the vibration energy transmitted from the vibration table to the camera support is attenuated by ≥ 40 dB. The air-bearing vibration isolation platform (natural frequency ≤ 2 Hz, attenuation ≥ 40 dB) effectively isolates the interference of the vibration table.

[0058] The innovative Kalman filter fusion strategy combines the global rigid body displacement information of the reference points and the local deformation details of DIC, realizing the reconstruction of the full-field strain field with high spatio-temporal resolution and meeting the dual requirements of accuracy and real-time performance for dynamic testing.

[0059] Furthermore, the vibration synchronization control interface can open the light source exposure window at the peak acceleration moment of the vibration table, and the deviation between the camera exposure time and the light source pulse width ≤ 0.1 μs.

[0060] Further explanation is that it also includes an online visualization terminal and an early warning module. The online visualization terminal real-time displays a deformation heat map, a strain evolution curve, and a displacement-time history diagram of key positions. The early warning module triggers an alarm signal when the local strain exceeds the material yield threshold.

[0061] Further explanation is that the wavelength of the light source is in the near-infrared band of 700–900 nm, and the passband bandwidth of the optical filter is ≤10 nm.

[0062] In summary, compared with the traditional strain gauge measurement method, this embodiment provides a non-contact, full-field, high-frequency dynamic measurement capability for the system; compared with the existing DIC system, the dynamic measurement accuracy is significantly improved through motion blur elimination and data fusion technologies.

[0063] For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or basic characteristics of the present invention. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be included in the present invention. Any reference signs in the claims should not be regarded as limiting the claims involved.

Claims

1. A real-time deformation monitoring system for a vibration table test sample based on machine vision, characterized in that, Including: Active modulation light source module: Adopts a modifiable light source with high frequency and narrow pulse width, whose light emission timing is synchronized with the motion phase of the vibration table, and is used to eliminate motion blur during sample vibration; Multi-camera high-speed acquisition subsystem: At least two high-frame-rate industrial cameras are fixed on a rigid bracket at orthogonal angles. The bracket is independently installed from the vibration table through a vibration isolation base. Each camera is equipped with an optical filter to match the light source wavelength and is directly connected to the processing unit through a gigabit network port; Real-time deformation analysis unit: Built-in with a reference point dynamic tracking module, the reference point dynamic tracking module identifies the high-contrast coded reference points pre-placed on the sample surface and resolves their 6-DOF displacements in the vibration coordinate system in real time; Full-field strain field reconstruction module: Performs sub-pixel level displacement vector calculation on the speckles on the sample surface based on the digital image correlation method (DIC); Motion compensation fusion module: Uses the reference point displacement data to perform rigid body motion compensation on the DIC result and outputs a pure deformation strain field; Vibration synchronization control interface: Receives the phase trigger signal from the vibration table controller and dynamically adjusts the light source exposure window and the camera acquisition timing; 2. The real-time deformation monitoring system for the test sample of the vibrating table based on machine vision according to claim 1, wherein: The active modulation light source module includes an LED array with a pulse width ≤ 1 μs, and the drive circuit of the LED array responds to the synchronization signal sent by the vibration synchronization control interface; 3. The real-time deformation monitoring system for the test sample of the vibration table based on machine vision according to claim 1, characterized in that: The reference points are dot matrices made of rare earth fluorescent materials, whose excitation wavelength matches that of the active modulation light source. Each reference point has a unique binary coding pattern, and the spatial resolution ≥ 0.05 mm; 4. The real-time deformation monitoring system for the vibration table test sample based on machine vision according to claim 1, characterized in that: The real-time deformation analysis unit further includes a multi-scale feature fusion module. The multi-scale feature fusion module performs Kalman filter fusion on the rigid body displacement data tracked by the reference points and the local strain data of DIC in the space-time domain, outputs a full-resolution strain field, and the update frequency ≥ 500 Hz; 5. The real-time deformation monitoring system for the vibration table test sample based on machine vision according to claim 1, wherein: The motion compensation fusion module performs the following operations: (a) Establish a local coordinate system with the reference point group as the reference; (b) Map the DIC displacement vector to this local coordinate system through affine transformation; (c) Eliminate the rigid body motion component and retain the elastic deformation component; 6. The real-time deformation monitoring system for the vibration table test sample based on machine vision according to claim 1, wherein: The vibration isolation base adopts an air-bearing vibration isolation platform, and the natural frequency of the air-bearing vibration isolation platform ≤ 2 Hz, and the vibration energy transmitted from the vibration table to the camera bracket is attenuated by ≥ 40 dB; 7. The real-time deformation monitoring system for a vibration table test sample based on machine vision according to claim 1, wherein: The vibration synchronization control interface can open the light source exposure window and the camera exposure time at the peak moment of the vibration table acceleration, and the deviation between the light source pulse width ≤ 0.1 μs; 8. The real-time deformation monitoring system for the test sample of the vibration table based on machine vision according to claim 1, wherein: It also includes an online visualization terminal and an early warning module. The online visualization terminal displays the deformation heat map, strain evolution curve and displacement-time history map of key positions in real time. The early warning module triggers an alarm signal when the local strain exceeds the material yield threshold; 9. The real-time deformation monitoring system for the test sample of the vibration table based on machine vision according to claim 1, characterized in that: The wavelength of the light source is in the near-infrared band of 700–900 nm, and the passband bandwidth of the optical filter ≤ 10 nm;

Images