Free space sound field measurement method and system

Abstract

The invention belongs to the technical field of acoustic measurement, and discloses a free space sound field measurement method and system, and the method comprises the steps: constructing a free light path sound field measurement system based on a Michelson interferometer principle; acquiring a reference image, introducing a sound field to be measured, and controlling a loudspeaker or a laser beam to perform translation and rotary scanning through a computer to complete interference pattern data acquisition; preprocessing the acquired interference pattern, identifying the edge of the interference fringes by using an edge detection technology, and carrying out light phase variation calculation; and performing sound field reconstruction by adopting a Radon inverse transformation algorithm, converting sound field integral information carried by a laser beam into a sound pressure value of each point in a real space, and reconstructing a free field sound pressure distribution diagram. The method has the advantages of no damage and low cost, the measurement precision of the sound field and the integrity and reliability of reconstruction are remarkably improved, the nano-scale optical path change caused by the sound field can be effectively captured, and the high precision of sound field measurement and the accuracy of sound source recognition are ensured.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of acoustic measurement, and particularly relates to a free space sound field measurement method and system, and particularly relates to a new technical method and system for realizing non-destructive, high-precision and non-invasive free space sound field measurement by using an acousto-optic effect. BACKGROUND

[0002] The existing acousto-optic effect measurement method has limitations for sound field measurement: the existing measurement method is mostly used to derive the sound field through indirect parameters or vibration information, and the demand for directly obtaining the spatial distribution of sound pressure of the sound field still needs to be met. In addition, the sound field measurement method based on the laser beam deflection effect obtains sound field information by sensing the deflection displacement of the laser beam passing through the sound field through a position sensitive detector (PSD). However, this method usually requires a strong sound field intensity, especially in the ultrasonic frequency band, to produce a sufficient and measurable laser beam deflection. Research shows that it can only obtain a stable deflection signal when measuring an ultrasonic sound pressure of the order of 20 kPa. This means that for weak sound fields or low-frequency sound fields audible to the human ear, the sensitivity of the laser beam deflection effect may not be sufficient for accurate measurement. Mach-Zehnder interferometers (MZI) have also been used for optical measurement of sound field information. Although MZI has an advantage in quantitative measurement of phase values and realizes two-dimensional data acquisition through a CCD camera, its optical system is usually more complex and contains more beam splitters and mirrors, which increases the difficulty and cost of system construction and later calibration. In addition, when the sound field frequency is high, the MZI combined with the CCD camera can only work as an average light intensity detector, and it is difficult to directly obtain the instantaneous phase information. Although the existence of the sound field can be confirmed through differential image processing, the general method for accurately inverting the instantaneous phase deviation from the average light intensity has not been clearly defined, and its calibration and verification of the actual sound pressure distribution still need further research. These factors together limit the wide application of the existing acousto-optic effect measurement technology.

[0003] Through the above analysis, the problems and defects of the prior art are: (1) Traditional sound field measurement usually uses multiple sound sensors, and the traditional sensor array measurement method has inherent invasiveness problems, limited measurement accuracy, high cost and complex operation.

[0004] (2) The existing sound field measurement method based on the acousto-optic effect is costly, and the optical system is usually more complex, which will increase the difficulty and cost of system construction and later calibration, and has limitations for sound field measurement. SUMMARY

[0005] To overcome the problems in the related art, the present application discloses a free space sound field measurement method and system, and the technical solution is as follows: The application is implemented by a free space sound field measurement method, comprising the following steps: S1, light path building: a free light path sound field measurement system based on the principle of Michelson interferometer is constructed; S2, sound field data acquisition: a reference image is acquired, a sound field to be measured is introduced, a loudspeaker or a laser beam is controlled by a computer to perform translation and rotation scanning, and interference pattern data acquisition is completed; S3, interference fringe image processing and light phase information extraction: the collected interference pattern is preprocessed, an edge detection technology is used to identify the edge of the interference fringe, and the light phase change amount is calculated; S4, sound field sound pressure reconstruction: a Radon inverse transform algorithm is used for sound field reconstruction, sound field integral information carried by the laser beam is converted into a sound pressure value of each point in the real space, and a free field sound pressure distribution map is reconstructed.

[0006] In step S1, the light path building comprises: The free light path sound field measurement system uses a continuous laser as a light source, and the laser beam emitted by the continuous laser is expanded by an expansion mirror; the laser beam after expansion is incident on a beam splitter and is divided into measurement light and reference light; The measurement light passes through a region where the sound field to be measured is placed, is incident on a mirror M1, and is reflected back to the beam splitter; the reference light is incident on a mirror M2 and is reflected back to the beam splitter; The two-way reflected light recombines at the beam splitter, undergoes equal-thickness interference, and forms interference fringes with light and dark alternation; the interference fringes are enlarged by an objective lens and recorded by a CCD camera.

[0007] Further, the sound field to be measured is generated by a loudspeaker, and all optical elements are accurately fixed and adjusted by a high-stability series adjustment frame. All optical elements (laser, beam splitter, expansion mirror, mirror, CCD, etc.) are installed on a high-stability series adjustment frame and rigidly fixed on an optical vibration isolation platform.

[0008] The adjustment frame is provided with a fine adjustment screw mechanism, which can finely adjust the pitch angle, horizontal angle and height to ensure that the light beam propagates along the central axis of the platform.

[0009] After fixing, the adjustment frame locking screw is tightened to avoid angle deviation of the elements due to slight vibration.

[0010] After the laser is fixed, the pitch angle and horizontal position are adjusted to make the light beam propagate along the center of the optical platform.

[0011] After the expansion mirror is installed, the height is accurately adjusted by the adjustment frame to make the expanded light spot uniformly cover the center of the beam splitter.

[0012] The beam splitter is fixed on the adjusting frame, the position is controlled about 10-15 cm behind the beam expander, and the reflected light and the transmitted light are made to be perpendicular to the center of the mirrors M1 and M2 respectively by adjusting the height and angle.

[0013] The height and angle of M1 and M2 are adjusted so that the reflected light coincides with the incident light when returning, and clear interference fringes are formed at the beam splitter.

[0014] In step S2, the sound field data acquisition includes: Acquiring reference image: the initial interference fringe pattern is photographed and recorded by the CCD camera without sound field disturbance, as the reference image for subsequent data processing; Introducing sound field: introducing the sound field to be measured into the measurement light path, so that the sound field disturbance passes through the measurement light in this area; Scanning measurement: the loudspeaker or laser beam is controlled by the computer to perform translational and rotational scanning; The translational-scanning and rotational-scanning processes are repeated to complete the interference pattern data acquisition at multiple positions and angles.

[0015] Further, the scanning measurement includes: fixing the position of the laser beam, and the loudspeaker performs step-by-step translation and rotation in a two-dimensional plane; the loudspeaker is fixed on a stepping motor, and the loudspeaker is moved by the motor; after each translation by a certain distance or rotation by a certain angle, the interference fringe pattern disturbed by the sound field is photographed and recorded in real time by the CCD camera.

[0016] In step S3, the interference fringe image processing and light phase information extraction include: (1) Image preprocessing: performing filtering and denoising preprocessing operation on the collected interference pattern; (2) Fringe edge detection: using edge detection algorithm to identify the edge of the interference fringe; (2) Light phase change calculation: according to the linear relationship between the movement of the interference fringe and the optical path difference, and the relationship between the change of the optical path difference and the light phase change, the fringe movement is converted into the light phase change caused by the sound field when the light beam passes through the sound field.

[0017] Further, the edge detection algorithm locates the edge information in the image by detecting the mutation area of the pixel gray value, and the mutation area includes the object contour and texture change; the sub-pixel edge detection algorithm is used to analyze the fringe displacement, and the actual movement distance of the interference fringe is obtained; the four-stage edge detection algorithm of Gaussian filtering, Sobel operator gradient calculation, non-maximum suppression and double-threshold segmentation is suitable for the sub-pixel level displacement measurement requirement of the interference fringe, and the specific process is as follows: For the random noise such as light intensity fluctuation and electronic noise in the interference fringe image collected by the CCD, the image is smoothed by using two-dimensional Gaussian function, and the expression is: ; ; wherein, is a two-dimensional Gaussian function, a kernel function for smoothing the image, is an exponential function with base of natural constant e , which is a core component of the two-dimensional Gaussian function, used to describe the decay property of the function value with x , y , is a coordinate variable of the image in the horizontal direction, used to determine the position-dependent value of the two-dimensional Gaussian function in the horizontal direction, is a coordinate variable of the image in the vertical direction, used to determine the position-dependent value of the two-dimensional Gaussian function in the vertical direction, is the standard deviation of the Gaussian kernel, , used to balance the noise suppression and edge blurring; is the gray value of the filtered image, is the gray value of the original interference fringe image, is a convolution operation; the gradient of the filtered image in the direction is calculated by the Sobel operator to determine the amplitude and direction of the edge, expressed as: ; ; ; ; wherein, are the directional gradients, is the gradient amplitude, used to reflect the edge intensity; is the gradient direction, used to reflect the edge orientation and match the vertical direction of the interference fringe; the local extremum judgment is performed on the gradient amplitude along the gradient direction, the edge center pixel is retained, and the non-edge pixel is removed, expressed as: ; wherein, is the image after non-maximum suppression, the edge width is refined from 3 to 5 pixels to 1 pixel, and the broken edges are connected by setting a high threshold and a low threshold, is the adjacent pixel in the gradient direction; the sub-pixel edge positioning is used to improve the accuracy, the sub-pixel level edge coordinate calculation is realized based on the interpolation method, the gray values of the 3×3 neighborhood near the edge are fitted by a quadratic polynomial, and the gray values near the edge are set to satisfy , and Derivation and let the derivative be 0, get sub-pixel coordinates: ; In the formula, is a polynomial fitting coefficient, is a sub-pixel level edge coordinate, and the positioning accuracy is 0.1 pixels; is a pixel coordinate.

[0018] Further, the actual displacement distance of each interference fringe in the horizontal or vertical direction is calculated as: The interference fringe spacing of the Michelson interferometer in the case of equal thickness interference is: ; In the formula, is the interference fringe spacing, is the laser wavelength, is the mirror angle; The fringe movement amount is: ; In the formula, is the optical path difference changed when the light path passes through the sound field; The actual movement distance of the fringe is: ; In the formula, is the phase change amount; Through the linear relationship between the phase change amount and the actual movement distance of the fringe, the sound field is inversely calculated.

[0019] In step S4, the sound field sound pressure reconstruction includes: inputting the obtained light phase change amount data under different projection angles and positions into the Radon inverse transform algorithm for sound field reconstruction to generate a free field sound pressure distribution map; The reference system of the sound field is coordinate system, the coordinate system after rotating 0 angle becomes coordinate system, The transformation relationship between the coordinate system and coordinate system is: ; The projection integral process is: ; In the formula, is the sound pressure to be measured in the sound field, is the sound pressure integral along the direction path , and is the path of the light passing through the sound pressure disturbance area; According to the dynamic offset, the relationship between the Radon transform of the sound pressure in the time domain and the sound pressure is: ; In the formula, is the wave number of light propagation in vacuum, is the pressure optical coefficient of the medium, is the phase measured by the interferometer, is the isentropic pressure optical coefficient, is the integral of the sound pressure along the path in the direction of , is the sound pressure of the reconstructed sound field at coordinates and time ; ; In the formula, is the relevant refractive index parameter of the medium, is the coefficient related to the characteristics of the medium, is the reference pressure; The measured sound field pressure is reconstructed by Radon inverse transform as: ; In the formula, is the sound pressure of the reconstructed sound field at coordinates and time , is the Radon inverse transform operator, which is used to deduce the sound pressure distribution from the projection integral result.

[0020] Another object of the present application is to provide a free space sound field measurement system for regulating the free space sound field measurement method, which comprises: A laser for emitting a laser beam with good coherence, preferably a continuous laser; A beam expander for expanding the laser beam; A beam splitter for splitting the expanded laser beam into measurement light and reference light; A mirror for reflecting the measurement light and the reference light to cause re-converging interference; wherein the mirror M1 is placed at the end of the region to be measured, so that the measurement light passes through the sound field and is reflected; A loudspeaker / sound source placed in the measurement light path as a sound field to be measured, for generating sound wave disturbance; A scientific grade CCD camera placed in the interference fringe forming area for recording interference patterns; An objective lens for magnifying the interference fringes and improving the image resolution recorded by the CCD camera; A computer and control system for controlling the scanning motion of a loudspeaker / sound source, data acquisition of a CCD camera, and performing image processing and sound field reconstruction algorithms. High-stability series of adjusting frames for accurately fixing and adjusting the position of optical elements In combination with all the technical solutions described above, the present application has the following beneficial effects: First, to address the shortcomings of existing sound-light effect measurement methods in obtaining accurate instantaneous phase information and complete sound pressure field reconstruction, the present application realizes nanometer-level high-precision measurement of the movement amount of interference fringes by combining optical imaging amplification and sub-pixel edge detection algorithms, thereby directly and accurately obtaining light phase change information. The present application uses the Radon inverse transform algorithm to reconstruct the complete free-field sound pressure spatial distribution from high-precision light phase data, and the feasibility has been verified through simulation. Compared with existing methods, the present application not only has the advantages of non-destructive and low cost, but also significantly improves the accuracy of sound field measurement and the completeness and reliability of reconstruction.

[0021] The sound-light effect-based free-space sound field measurement system and method proposed by the present application has the following significant advantages compared with existing technologies: 1. Non-destructive and non-invasive measurement: The present application uses a laser beam to pass through the sound field and perceives sound field information through the sound-light effect. The laser itself does not cause physical disturbance to the sound field, realizing truly non-contact and non-interference measurement. This solves the problem of the physical disturbance caused by the sensor body in the traditional microphone array measurement.

[0022] 2. High measurement accuracy: By using the equal-thickness interference principle of the Michelson interferometer, combined with optical imaging amplification and sub-pixel edge detection algorithms, sub-nanometer-level displacement resolution can be achieved in theory. This enables the present application to accurately perceive nanometer-level optical path changes caused by the sound field, thereby realizing high-precision measurement of the sound field.

[0023] 3. High cost-effectiveness: Compared with existing sound-light effect measurement equipment such as expensive laser Doppler vibrometers on the market, the present application only needs necessary continuous lasers, high-precision optical lenses (such as mirrors and beam splitters) and CCD cameras to build a measurement system, greatly reducing the measurement cost.

[0024] 4. Directly obtaining sound pressure distribution of sound field: The present application can directly reconstruct the sound pressure spatial distribution of the sound field from the measured light phase projection information by using the Radon inverse transform algorithm, providing more intuitive and comprehensive data for sound field analysis.

[0025] 5. Wide applicability: The present method is suitable for sound field measurement in free space and can be applied to various scenarios that require accurate acquisition of sound field information.

[0026] Second, the purpose of the present application is to propose and implement a non-invasive, high-precision, low-cost free-field sound field measurement system and method to address the invasiveness and precision limitations of traditional sound field measurement methods (such as sensor array method) and the problem of expensive existing sound-light effect measurement method equipment. The present application builds a free optical path measurement system based on Michelson interferometer, uses the principle of sound-light effect, combines advanced image processing and signal processing algorithms, accurately obtains sound field information, and realizes the reconstruction of sound field. Specifically, the measurement system designed by the present application can theoretically achieve sub-nanometer displacement resolution, such as under the condition of equal thickness interference inclination angle of 0.1° and laser wavelength of 532nm, the displacement sensitivity can reach about 0.17nm / pixel, and theoretically sub-nanometer displacement resolution can be achieved. This high-precision detection capability can effectively capture the nanometer-level optical path change caused by the sound field, thereby ensuring the high precision of the sound field measurement and the accuracy of the sound source identification.

[0027] Third, traditional sound field measurement methods, such as sensor array method, have inherent limitations, such as the need to place sensors directly in the sound field, thereby interfering with the sound field, and high cost and complex operation. While existing sound-light effect measurement devices on the market, such as laser Doppler vibration meters, can achieve non-invasive measurement, their high price greatly limits their widespread application. The present application builds a system based on Michelson interferometer, which only needs to use necessary continuous lasers, high-precision optical lenses and CCD cameras and other relatively low-cost devices. This low-cost advantage makes it more attractive in scientific research and industrial fields, and can provide a cost-effective alternative for various enterprises and research institutions that need accurate sound field measurement.

[0028] In addition, the present application can directly reconstruct the sound pressure spatial distribution of the sound field, providing more intuitive and comprehensive data, which has important commercial application value in noise control, product acoustic quality testing (such as household appliances, automobiles, etc.) and sound stealth performance evaluation in military field. Its non-invasive, high-precision and high-reliability can meet the strict requirements of sound field measurement scenarios, thereby expanding new market opportunities.

[0029] Fourth, the present application realizes non-invasive, high-precision sound field reconstruction: in the prior art, sound-light effect-based measurement methods mostly derive sound field through indirect parameters or vibration information, making it difficult to directly obtain accurate sound pressure spatial distribution of sound field. This invention uses Radon inverse transform algorithm to directly reconstruct complete sound pressure distribution from light phase change data (i.e. sound field integral information) obtained from different angles and positions.

[0030] Solve the limitations of existing acousto-optic effect methods: Some methods based on Mach-Zehnder interferometer (MZI) also use acousto-optic effect, but their optical systems are complex, difficult to calibrate, and it is difficult to directly obtain instantaneous phase information when dealing with high-frequency sound field. The Michelson interferometer used in the present application combines optical amplification and sub-pixel edge detection algorithm, which can realize nanoscale optical path change measurement, and solves the shortcomings of traditional methods in obtaining accurate instantaneous phase information and complete sound pressure field reconstruction.

[0031] Fifth, for the traditional sensor array method, the sensor itself will cause sound scattering, causing disturbance to the original sound field, thereby affecting the measurement accuracy and spatial resolution. The present application uses laser beam for non-contact measurement, which fundamentally solves the inherent invasiveness problem and realizes truly non-destructive measurement.

[0032] Cost and accuracy balance problem: Although existing devices such as laser Doppler vibrometer can achieve high-precision measurement, the high cost limits its application. The present application significantly reduces the cost of the device while ensuring high measurement accuracy, providing a feasible solution that balances performance and economy for the industry.

[0033] The present application proves that by optimizing a relatively simple optical system (Michelson interferometer) and combining innovative data processing algorithms (sub-pixel edge detection and Radon inverse transform), performance comparable to or even better than expensive equipment can be achieved. This method opens up a new way for acoustic measurement and promotes the technological progress in this field. BRIEF DESCRIPTION OF DRAWINGS

[0034] The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure; Figure 1 is a flow chart of the free space sound field measurement method provided by the embodiment of the present application; Figure 2 is a principle diagram of the free space sound field measurement method provided by the embodiment of the present application; Figure 3 is an experimental optical path diagram provided by the embodiment of the present application; Figure 4 is a sound field data acquisition experiment process diagram provided by the embodiment of the present application; wherein (a) is a translation scan, and (b) is a rotation scan; Figure 5 is an actual fringe pattern obtained by using a CCD provided by the embodiment of the present application; Figure 6The simulation diagram of interference fringe change of the Michelson interferometer when being disturbed by a sound field is provided by the embodiment of the present application; wherein, (a) is a sound pressure distribution of a simulation point sound source, (b) is a refractive index change after being affected by a sound field, (c) is a phase change diagram of a light beam transversely incident through a sound field disturbance at different incident positions, and (d) is a contrast diagram of fringe movement after interference of a light beam passing through a sound field center position. Figure 7 The schematic diagram of the Radon transform principle is provided by the embodiment of the present application. In the figure, Laser is a laser, BS is a beam splitter, LBE is a beam expander, L is a lens, M1 and M2 are mirrors, and CCD is an industrial camera. DETAILED DESCRIPTION

[0035] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the specific embodiments of the present application are described in detail below in combination with the drawings. In the following description, a large number of specific details are set forth in order to facilitate a full understanding of the present application. However, the present application can be implemented in many other ways different from those described herein, and those skilled in the art can make similar improvements without departing from the spirit of the present application, so the present application is not limited to the specific embodiments disclosed below.

[0036] The innovation point of the present application is: 1. The present application proposes and realizes a free optical path sound field nondestructive measurement method based on the Michelson interferometer: innovatively using the principle of acousto-optic effect, building a Michelson interferometer to realize laser tomographic scanning, so as to realize non-contact and non-interference measurement of the sound field in the free optical path, solving the problem of sound field disturbance of the traditional contact sensor.

[0037] 2. The present application realizes high-precision measurement of the sound field by improving the Michelson interferometer and combining advanced data processing algorithms: the present application combines optical imaging magnification technology and sub-pixel edge detection algorithm to finely process the measurement results of the Michelson interferometer, so as to have high sensitivity to the small light phase change caused by the sound field, and theoretically realize sub-nanometer level displacement resolution, thereby ensuring the high precision of the sound field measurement.

[0038] The principle of the free space sound field measurement method and system provided by the embodiment of the present application is: Sound propagation essentially involves changes in pressure in the medium of propagation. When a sound field exists in a medium (e.g. air), the changes in sound pressure cause changes in the optical properties of the medium, i.e. the refractive index changes. According to the Gladstone-Dale relation and Taylor expansion, the change in refractive index of the medium is directly proportional to the sound pressure under weak acousto-optic interaction. The phase of light is related to the propagation distance of light in the medium and the refractive index of the medium. Therefore, the change in refractive index of the medium caused by the sound field will cause the phase of the light beam passing through the sound field to change. By measuring the amount of change in the phase of light, the sound pressure of the sound field can be indirectly calculated in theory.

[0039] Sound propagation essentially involves changes in pressure in the medium of propagation. Therefore, sound information can be extracted from the sound pressure fluctuations caused by the sound field. Since the refractive index of air changes due to the influence of the sound field, and the phase of light also changes, as long as the amount of change in the phase of light is measured, the sound pressure of the sound field can be indirectly calculated in theory. Based on this, the change in the refractive index of air caused by sound vibration is captured using laser, thereby extracting the sound pressure information therein.

[0040] The propagation of sound changes the density and pressure of the medium, thereby affecting the refractive index of air. Assuming that this process is adiabatic, the relationship between the pressure and density in air is given by: (1) where and are the pressure and density under static conditions, respectively, and is the specific heat ratio.

[0041] When sound propagates, corresponds to the superposition of static pressure and sound pressure, i.e. The effect of density change on light propagation can be determined by combining the mechanical and optical properties of the medium. In 1863, Gladstone and Dale established an empirical relation between the refractive index and the density of various liquids: (2) where is the Gladstone-Dale constant.

[0042] Combining the static refractive index and the static density of air, we can derive: (3) Combining the above equation, we can derive the relationship between the pressure and the refractive index in air: (4) where, n0is the refractive index under standard atmospheric conditions.

[0043] When the sound pressure is much smaller than the static pressure , the expression can be approximated by a first order Taylor expansion: (5) Therefore, under weak acousto-optic interaction, the change in refractive index is proportional to the sound pressure. This indicates a linear relationship between the sound pressure and the refractive index of air. Light can be used to sense this change in refractive index caused by the sound field in air, and thus makes it possible to measure the sound pressure.

[0044] In a homogeneous, static medium, i.e. the electromagnetic properties of the medium are independent of the spatial location and time, the wave equation for the electromagnetic field is: (6) When a sound field is present in the medium, it is no longer homogeneous and isotropic. However, according to the research by P. M. Morse, et al., the wave equation in the above equation is still valid when the following constraint is satisfied:

[0045] (7) In the audible range of human ears, the constraint in the above equation can be fully satisfied. Therefore, the solution to the wave equation is: (8) where, is a complex number, which contains the initial amplitude and phase information of the light; and are the angular frequency and phase of the light, respectively.

[0046] The relationship between the phase and the refractive index is: (9) where, is the wave number of light propagating in vacuum, is the distance of light propagating in air, i.e. the length of the optical path.

[0047] In air, the relationship between the phase of light and the sound pressure in air can be obtained by combining the above equations: (10) where, is the shift in the phase of light in the absence of a sound field, which depends on the path length of light in the medium. This term can be regarded as a static shift ; The part of the phase of light that changes due to the change in the refractive index of air caused by the sound field is referred to as the dynamic shift : (11) The above equation shows the relationship between the path integral of the sound pressure field and the light phase. Therefore, as long as the change in the light phase is measured, the sound pressure field can be indirectly calculated in theory.

[0048] This invention utilizes the principle of equal-thickness interference in a Michelson interferometer to sense minute changes in the phase of light. The Michelson interferometer uses a beam splitter to separate light emitted from a light source into a measurement beam and a reference beam. The measurement beam passes through the region of the sound field to be measured, and its optical path changes due to variations in the refractive index of the medium caused by the sound field. After the measurement beam and the reference beam re-merge, they interfere, forming alternating bright and dark interference fringes. When the optical path of the measurement beam changes, the interference fringes shift. By comparing the amount of shift in the interference fringes before and after the presence of the sound field, the change in the phase of light can be obtained.

[0049] There is a linear relationship between the path integral of sound pressure and the phase of light in a sound field. The path integral of sound pressure obtained through interferometry can be viewed as a projection of the sound field onto the direction of light propagation; this projection method is the mathematical Radon transform. Therefore, by performing an inverse Radon transform on the projection information from multiple angles or locations, the spatial distribution of sound pressure in the sound field can be reconstructed.

[0050] Example 1, such as Figure 1 As shown, the free-space sound field measurement method provided in this embodiment of the invention includes the following steps: S1, Optical path construction: Constructing a free-path acoustic field measurement system based on the principle of Michelson interferometer; The free-path acoustic field measurement system uses a continuous laser as its light source. The emitted laser beam is first expanded by a beam expander. The expanded laser beam is then incident on a beam splitter and split into a measurement beam and a reference beam. The measurement beam passes through the area where the acoustic field to be measured (e.g., generated by a loudspeaker) is placed, strikes mirror M1, and is reflected back to the beam splitter. The reference beam strikes mirror M2 and is also reflected back to the beam splitter. The two reflected beams re-merge at the beam splitter, resulting in equal-thickness interference and forming alternating bright and dark interference fringes. These interference fringes are magnified by an objective lens and recorded by a scientific-grade CCD camera. All optical components are precisely fixed and adjusted using a high-stability series of adjustment frames.

[0051] Traditional Mach-Zehnder interferometers have complex optical systems, making calibration difficult and costly. The Michelson interferometer used in this invention has a relatively simple structure, and all optical components are precisely fixed and adjusted using a highly stable series of adjustment frames, effectively reducing the difficulty of system setup and calibration. This results in a low-cost, highly stable measurement system that effectively reduces the cumulative impact of environmental disturbances (such as vibration and temperature) on optical path difference.

[0052] The laboratory Michelson interferometer uses a continuous laser with a wavelength of 532 nm, a non-polarizing flat beam splitter with a diameter of 25.4 mm is selected as the beam splitter, the thickness is 5.0 mm, the splitting ratio is 50:50 (R:T), the anti-reflection film is 400-700 nm, the beam expander is a variable magnification laser beam expander with a working wavelength of 532 nm and a magnification of 2-10, the mirror is a high-power laser mirror with a diameter of 12.7 mm, a working wavelength of 532 nm and 1064 nm, and a surface flatness of λ / 10. The optical path calibration first fixes the laser, adjusts the pitch angle and horizontal position of the laser, and makes the laser beam propagate along the central axis of the optical platform; place the beam expander at the laser light outlet, adjust the height of the beam expander, and make the expanded parallel light spot evenly cover the center of the subsequent beam splitter; fix the beam splitter (BS) on the adjustment frame and place it 10-15 cm behind the beam expander; fine-tune the height and angle of the beam splitter so that the reflected light (reference light path) is perpendicular to the center of M1 mirror, and the transmitted light (measurement light path) is perpendicular to the center of M2 mirror; adjust the height of the adjustment frame of M1 and M2 so that the centers of the reflecting surfaces of the two mirrors are consistent with the height of the incident light beam. Adjust the fine-tune knob of M1 so that the reference light reflected by M1 returns along the original light path, and finally coincides with the incident light spot near the laser light outlet; adjust the fine-tune knob of M2 so that the measurement light returns along the original light path after being reflected by M2, and converges with the reference light at the beam splitter; at this time, place a CCD camera behind the beam splitter to observe the blurred interference fringes, and then fine-tune the mirror surface angle of M2 and M1 to obtain bright and clear interference fringes. The experiment ensures the stability of the interference from four dimensions of hardware vibration isolation, environmental control, optical path optimization and data assistance. First, the Michelson interferometer is built on an optical vibration isolation platform, and all optical elements (laser, mirror, CCD) are rigidly fixed on the platform through a special clamp to avoid displacement of the elements themselves. After the optical path calibration is completed, tighten the locking screw of the MT-AM1T adjustment frame to prevent the mirror from being angle offset due to slight vibration; then use a black cloth to build a closed windproof cover with a size of 1200x1000x1000mm, which wraps the entire interference light path inside to avoid air refractive index fluctuations in the measurement light path caused by laboratory air flow (such as air conditioner wind, personnel breathing). The experimental area maintains a constant temperature through the air conditioner to avoid thermal expansion and contraction of optical elements due to temperature changes; the experiment reduces the accumulated error of the optical path difference by shortening the optical path length, and the total length of the reference light path and the measurement light path is controlled within 1-1.5m to reduce the cumulative influence of environmental disturbances (such as vibration, temperature) on the optical path difference; finally, in terms of data acquisition, the Matlab control framework is used to realize the synchronization of loudspeaker driving and CCD acquisition, and the CCD only collects images within 100ms after each loudspeaker moves or rotates, which shortens the collection time and reduces the drift of the fringes. At the same time, multiple acquisitions are performed during the collection process, and the interference fringes under the same sound field conditions are collected for 10 times in succession, and the image averaging algorithm is used to eliminate the fringe jitter caused by random vibration.

[0053] S2, sound field data acquisition: collect reference image, introduce the sound field to be measured, and complete the data acquisition of interference pattern by computer control of loudspeaker or laser beam translation and rotation scanning; (1) Collect reference image: in the absence of sound field interference, the initial interference fringe pattern is recorded by CCD camera as the reference image for subsequent data processing; (2) Introduce sound field: introduce the sound field to be measured in the measurement light path, so that the sound field disturbance passes through the measurement light in this area; (3) Scan measurement: translation and rotation scanning of loudspeaker or laser beam are controlled by computer; For example, the laser beam position is fixed, and the loudspeaker is stepwise translated and rotated in a two-dimensional plane. The loudspeaker can be tightly fixed on the stepping motor, which drives it to move. After each translation by a certain distance or rotation by a certain angle, the interference fringe pattern disturbed by the sound field is recorded in real time by the CCD camera.

[0054] In the scanning measurement, the loudspeaker is placed on the motor turntable, and the turntable is moved vertically to the beam of the Michelson interferometer measurement arm to realize parallel line scanning of the sound field. The moving range is 30 cm on the left and 30 cm on the right of the measurement arm, and the scanning angle range is 0°-180°. Because the projection data of θ≥180° in the parallel beam scanning mode is a copy of the projection of θ-180°, data redundancy can be effectively avoided and the measurement process can be simplified. The quality of two-dimensional reconstruction of the sound field depends on the spatial and angular resolution of the scanning. If the resolution is too low, the reconstructed sound field will be unclear and details will be lost. If the resolution is too high, the reconstruction process will be extremely slow and the efficiency will be reduced. Therefore, the reconstruction of the sound field needs to be balanced between the quality and efficiency of the reconstruction. At the same time, the reconstruction resolution is limited by the laser diameter. If the spatial resolution is higher than the laser diameter, the laser cannot be considered as an ideal linear beam. Therefore, it is necessary to ensure that the translation distance is higher than the laser beam diameter. The experimental laser diameter is 1.5 mm, the spatial resolution (translation step) is set to 2 cm, and the angular resolution (rotation angle interval) is 10°. The wavelength of 2 kHz sound wave in air is about 17 cm, and the spatial resolution of 2 cm is less than 1 / 2 (8.5 cm) of it, which meets the Nyquist sampling theorem and avoids aliasing effect. The driving signal of the loudspeaker and the data acquisition system controlled by the Matlab program realize synchronous triggering of signal transmission and acquisition, ensuring that the sound field excitation and the light phase measurement of the interferometer are synchronized.

[0055] (4) Repeat the translation-scan and rotation-scan process to complete the interference pattern data acquisition at multiple positions and angles.

[0056] The traditional sensor array method needs a large number of sensors, is high in cost and complex in calibration. The present application can realize data acquisition by only one laser beam and sound source through scanning measurement, greatly reducing the hardware cost and operation complexity. The omnidirectional and multi-angle data acquisition of the sound field is realized, and sufficient projection information is provided for subsequent accurate reconstruction. In the scanning mode, the spatial resolution and angular resolution can be flexibly set according to the frequency of the sound wave. For example, the wavelength of a 2kHz sound wave in air is about 17cm, and a spatial resolution of 2cm is less than 1 / 2 (8.5cm) of it, which meets the Nyquist sampling theorem and can avoid aliasing effect.

[0057] S3, interference fringe image processing and light phase information extraction: the collected interference pattern is pretreated, the edge of the interference fringe is identified by using edge detection technology, and the light phase change is calculated; For each group of "sound field interference fringe pattern" and "no sound field interference fringe pattern" collected in step S2, the following processing is performed: 1. Image preprocessing: the collected interference pattern is pretreated, such as filtering and denoising, to improve the image quality.

[0058] 2. Fringe edge detection: the edge detection technology in image processing is used to identify the clear edge of the interference fringe. The algorithm locates the edge information in the image by detecting the mutation area of the pixel gray value (such as object contour, texture change, etc.). In order to improve the measurement accuracy, a sub-pixel edge detection algorithm can be used to analyze the fringe displacement and accurately obtain the actual movement distance of the interference fringe. The specific processing process is as follows: (1) Image acquisition and region cropping: first, the interference fringe image is acquired from the CCD camera, and the image is cropped by the region of interest (ROI) to focus on the core area of the interference fringe.

[0059] (2) Image filtering and gradient calculation: the cropped image is applied with an edge detection filter, which is a key step after acquisition. The filter is used to obtain the edge gradient and direction of the image. Common operators such as Canny operator used in the present application can be used for this purpose. In addition to this, Sobel operator, edges_image, derivate_gauss, edges_color, etc. can also be used for this purpose. These filters identify potential edges by finding parts of the image where the pixel gray value changes significantly, and after identifying the potential edges, the edge information of the fringe can be located through subsequent threshold processing.

[0060] (3) Edge extraction and threshold processing: After applying the filter, the high-light edges (i.e. the pixel points with high gradient) in the fringe image are extracted by threshold processing (such as threshold operator or hysteresis_threshold operator), and non-key edges are reduced to form edge thin lines.

[0061] (4) Non-maximum suppression and edge thinning: The extracted edges are further subjected to non-maximum suppression to ensure that the edge line width is single-pixel, and the accuracy of the edges is improved. Subsequently, the edges can be drawn or thinned using skelcton operator or the like. For discontinuous edges, merging, separating background and other processing are required to obtain a complete contour.

[0062] (5) Sub-pixel edge positioning: In order to improve the measurement accuracy, a sub-pixel edge detection algorithm is used. This algorithm realizes the sub-pixel level accurate positioning of the interference fringe edge by interpolating or curve fitting (for example, using Gauss-Seidel iteration method, Zernike moment method or interpolation method, etc.) on the gray information between pixels.

[0063] (6) Fringe displacement calculation: Obtain the sub-pixel level edge position of the reference image in the absence of sound field, and the sub-pixel level edge position of each scanning position and angle image in the presence of sound field. By comparing the two sets of edge data, the actual displacement distance of each interference fringe in the horizontal or vertical direction is accurately calculated.

[0064] The traditional image processing method has the problems of edge blur and low positioning accuracy in interference fringe measurement. The four-stage edge detection algorithm (Gaussian filter, Sobel operator, non-maximum suppression, and double threshold segmentation) and sub-pixel positioning technology are used to solve this problem. The edge positioning error is reduced from about 1.2 pixels of the traditional method to about 0.1 pixel, and the accuracy is improved by 10 times, and the processing efficiency is improved by 2 times. This high-precision detection capability can effectively capture the nanoscale optical path change caused by the sound field.

[0065] 3. Light phase change calculation: According to the linear relationship between the movement of the interference fringe and the optical path difference, and the relationship between the change of the optical path difference and the light phase change, the fringe movement is converted into the light phase change caused by the sound field when the light beam passes through the sound field.

[0066] The present application realizes the measurement of the sound field based on the equal-thickness interference of the Michelson interferometer. When the measurement light and the reference light interfere, vertical stripes of light and dark will be formed. With the change of the optical path of the measurement light, the interference stripes will move. By comparing the movement of the front and back stripes, the change of the optical path can be known in theory, and the change of the light phase can be obtained. Based on this, due to the existence of the sound field, the optical path of the measurement light will change, and the stripes will move when the equal-thickness interference with the reference light occurs. By comparing the movement of the stripes before and after the existence of the sound field, the change of the light phase can be known, and the measurement of the sound field can be realized.

[0067] The interference stripe spacing of the Michelson interferometer in the case of equal-thickness interference is: (12) In the formula, is the interference stripe spacing, is the laser wavelength, is the mirror angle; The movement of the stripes is: (13) In the formula, is the change of the optical path difference when the light path passes through the sound field; Therefore, the actual movement distance of the stripes is: (14) In the formula, is the phase change amount; Therefore, the phase change amount has a linear relationship with the actual movement distance of the stripes. After knowing the movement distance of the stripes, the change of the phase caused by the existence of the sound field can be known, and the sound field can be inversely calculated.

[0068] S4, sound field sound pressure reconstruction: the Radon inverse transform algorithm is used for sound field reconstruction, and the sound field integral information carried by the laser beam is converted into the sound pressure value of each point in the real space, and the free field sound pressure distribution map is reconstructed.

[0069] The obtained light phase change amount data (i.e. path integral information of the sound field) at different positions and angles is input into the sound field reconstruction algorithm. The present application uses the Radon inverse transform algorithm for sound field reconstruction. The Radon inverse transform can inverse the spatial distribution of the original two-dimensional sound field from the multi-angle line integral projection data. Through this algorithm, the sound field integral information carried by the laser beam can be converted into the sound pressure value of each point in the real space, so that the complete free field sound pressure distribution map can be reconstructed.

[0070] By using lasers to capture changes in the air's refractive index caused by sound vibrations, sound pressure information can be extracted. This data can then be used to reconstruct the sound pressure distribution of the sound field. The line integral of sound pressure can be viewed as a projection of the sound field onto the direction of light propagation; this projection method is known mathematically as the Radon transform.

[0071] The frame of reference for the sound field is The coordinate system, after rotating by 0 degrees, becomes Coordinate systems. There is a transformation relationship between these two coordinate systems: (15) This projection integration process can be described as: (16) In the formula, The sound pressure to be measured in the sound field. For along Directional path is The sound pressure integral, This represents the path of light through the region of sound pressure disturbance. Based on the relationship between phase and sound pressure in equation (11), the relationship between the Radon transform of sound pressure and sound pressure in the time domain can be obtained: (17) In the formula, Let be the wavenumber of light propagating in a vacuum. The photoluminescence coefficient of the medium, The phase measured by the interferometer. The isentropic pressure coefficient isotropic. The reconstructed sound field in coordinates place, time t The sound pressure level; Therefore, the sound pressure of the measured sound field can be reconstructed using the inverse Radon transform: (18) In the formula, The reconstructed sound field in coordinates place, time t sound pressure, This is the Radon inverse transform operator, which is used to infer the sound pressure distribution from the projection integral result.

[0072] The process of sound field reconstruction using the Radon inverse transform algorithm is as follows: Firstly, a series of light phase change amounts at different projection angles and projection positions are obtained from step S3. These light phase change amounts represent the line integral of the sound field (i.e. the refractive index change) along the propagation path of the laser beam, and can be regarded as the projection data of the sound field in the corresponding direction. The Radon transform of the sound pressure p of the sound field can be expressed as a line integral along a certain direction. Then, the collected multi-angle and multi-position projection data are taken as input, and a Radon inverse transform algorithm is applied. The Radon inverse transform is the inverse operation of the Radon transform, and its core idea is to reconstruct the original two-dimensional (or three-dimensional) function distribution according to the projection data obtained from multiple angles. In acoustic measurement, this means that the sound pressure value at each point on the path is inversely calculated from the sound pressure projection information, so as to realize the reconstruction of the sound field distribution. The output of the Radon inverse transform is the spatial distribution function of the original two-dimensional sound field . This means that the sound field information integrated along a certain path carried by the laser beam is successfully converted into the sound pressure value of each discrete point in the sound field space. Finally, the reconstructed two-dimensional sound pressure distribution data is visualized to generate a sound pressure distribution map.

[0073] Through the above process, the present application can convert the indirect projection information obtained by optical measurement into a complete free-field sound pressure spatial distribution map with physical meaning, and provide accurate data support for acoustic analysis and noise control.

[0074] In embodiment 2, the free-space sound field measurement system provided by the present application mainly includes: As shown in Figure 3 , the system uses a 532nm continuous laser as a light source. The laser beam emitted by the laser is used as the main light path, which first passes through the beam expander LBE. The expanded laser is incident on the beam splitter. The beam splitter is located at the center of the light path, which divides the incident light beam into two parts, one of which is incident on the M1 mirror as reference light and is reflected back to the beam splitter. The other is the measurement light which passes through the sound field to be measured and is finally incident on the mirror M2, which reflects it back to the beam splitter. The loudspeaker is placed in the measurement light path as the sound source to be measured, and is fixed on the guide rail. The guide rail is used to realize the translation and rotation scanning of the loudspeaker in the two-dimensional plane. The two beams reflected from M1 and M2 recombine at the beam splitter, and the interference fringes are formed. The interference fringes are enlarged by the objective lens, and then recorded by the CCD (scientific grade CCD camera) located below the objective lens. The computer is connected to the loudspeaker and the CCD through wires. The scanning motion of the loudspeaker is controlled, such as translation and rotation. The CCD camera is controlled to collect data, that is, to shoot interference patterns. The image data collected by the CCD are received and processed, and the image processing and sound field reconstruction algorithms are executed.

[0075] Laser: used to emit a laser beam with good coherence, such as a continuous laser.

[0076] Beam expander (LBE): Expands the laser beam to observe interference fringes.

[0077] Beam splitter (BS): Divides the expanded laser beam into two beams: measurement light and reference light.

[0078] Mirrors (M1, M2): Used to reflect measurement light and reference light to recombine and interfere. Mirror M1 is placed at the end of the sound field region to be measured, allowing the measurement light to pass through the sound field and reflect.

[0079] Loudspeaker / sound source: Placed in the measurement light path as the sound field to be measured, used to generate acoustic wave disturbances.

[0080] Scientific-grade CCD camera: Placed in the interference fringe formation area to record interference patterns.

[0081] Objective lens: Used to magnify interference fringes to improve image resolution recorded by the CCD camera.

[0082] Computer and control system: Used to control the scanning motion of the loudspeaker / sound source, data acquisition of the CCD camera, and perform image processing and sound field reconstruction algorithms.

[0083] High-stability series adjustment stand: Used to accurately fix and adjust the position of optical elements.

[0084] Example 3, Alternative to sound source movement: The movement of the loudspeaker can be achieved through various high-precision positioning devices such as robotic arms, automated rail systems, etc., rather than being limited to stepper motor drives. The scanning of the laser beam can also be achieved through optical scanning equipment such as high-speed scanning galvanometers, thereby keeping the sound source stationary.

[0085] Example 4, Alternative to CCD camera and objective lens: CMOS cameras or other photoelectric detectors with high resolution and high frame rate combined with high magnification microscope objectives can also replace the combination of scientific-grade CCD cameras and objective lenses, as long as they can meet the accuracy and resolution requirements of interference fringe image acquisition.

[0086] Example 5, Alternative to edge detection algorithm: In addition to Halcon-related operators and sub-pixel edge detection algorithms, other advanced image processing algorithms such as deep learning-based image segmentation and displacement estimation methods can also be used for interference fringe recognition and displacement extraction to further improve processing speed and accuracy.

[0087] Example 6, Alternative to interferometer type: Although the invention takes the Michelson interferometer as an example, other types of interferometers (such as Mach-Zehnder interferometers) can also be used as alternative solutions if they can achieve accurate measurement of the phase change of light caused by the sound field.

[0088] Example 7, Alternative to the laser type: As long as a laser beam with good coherence and appropriate wavelength can be provided, other types of lasers, such as pulsed lasers (if they can match the response speed of the interferometer and accurately measure the transient acoustic field), can also be used as alternatives to continuous lasers.

[0089] To further demonstrate the positive effects of the above examples, the present application is based on the above technical solutions to conduct the following experiments.

[0090] 1. Non-invasive measurement is achieved Traditional acoustic field measurement methods, such as sensor array method, have inherent invasiveness problem. The sensor body needs to be placed directly in the acoustic field to be measured, and its physical existence will cause acoustic scattering, thereby causing disturbance to the original acoustic field. This interference in the limited acoustic field space not only limits the number of acoustic sensors in the array, but also affects the measurement accuracy and spatial resolution.

[0091] The present application uses the principle of acousto-optic effect to sense the acoustic field signal by using only a laser beam to pass through the acoustic field. The laser itself does not cause physical disturbance to the acoustic field, achieving truly non-contact and non-interference measurement. This fundamentally solves the inherent problem of the influence of sensors on the acoustic field in traditional microphone array measurement.

[0092] 2. High measurement accuracy Sub-nanometer displacement resolution: The present application uses the equal-thickness interference principle of Michelson interferometer, combined with optical imaging magnification and advanced sub-pixel edge detection algorithm. This combination enables the system to accurately perceive the nanometer-level optical path change caused by the acoustic field. Under the condition of equal-thickness interference angle of 0.1° and laser wavelength of 532nm, the system can theoretically achieve a displacement sensitivity of about 0.17nm / pixel, realizing sub-nanometer displacement resolution.

[0093] Improved edge detection algorithm: The improved sub-pixel edge detection algorithm used in the present application reduces the fringe positioning error from about 1.2 pixels in traditional methods to about 0.1 pixel. Compared with traditional methods, its accuracy is improved by 10 times, and the processing efficiency is also improved by 2 times. This high-precision fringe displacement measurement ensures high accuracy of acoustic field measurement and accuracy of sound source identification.

[0094] Data acquisition stability: Through various measures such as hardware vibration isolation, environmental control (windshield and temperature constant), optical path optimization and data assistance (synchronous acquisition, multiple acquisition averaging), the present application effectively ensures the stability of the interference, reduces the fringe drift and jitter, and ensures the reliability of the measurement.

[0095] 3. Lower cost advantage: Compared with the existing acousto-optic effect measuring equipment such as the expensive laser Doppler vibrometer on the market, the application only needs necessary continuous lasers, high-precision optical lenses and relatively low-cost equipment such as scientific-grade CCD cameras, so that the measuring cost is greatly reduced.

[0096] 4. Theoretical and simulation verification feasibility The application has been verified by simulation. The simulation of a 1000Hz point sound source shows that it can cause nanoscale optical path changes, and the stripe movement is micron level, which needs high-resolution CCD cameras and edge detection algorithms for measurement. In addition, the Radon transform simulation of monopole and simply supported aluminum plate also verifies the feasibility of reconstructing the sound field by laser measurement data. These theories and simulation results together prove the great potential of the method in realizing high-precision, non-destructive sound field measurement.

[0097] The application aims at effective measurement of free optical path sound field information, and builds a free optical path measurement system based on Michelson interferometer, uses the characteristic that the disturbance of sound field to light causes the change of light phase, combines advanced image processing and signal reconstruction algorithm, and realizes non-destructive, high-precision and low-cost measurement of sound field.

[0098] The implementation flowchart of the method adopted by the application is shown in Figure 2 The laser emitted by the laser is expanded and then enters the Michelson interferometer, and is divided into measurement light and reference light. The measurement light passes through the sound field area to be measured and is combined with the reference light to produce interference fringes. The CCD camera captures the interference fringe images with and without sound field disturbance. Through image processing, including edge detection and sub-pixel displacement analysis, the movement amount of the interference fringes is accurately obtained, so that the light phase change information caused by the sound field is obtained. Finally, the sound pressure spatial distribution of the sound field is reconstructed from the light phase change data of multiple positions and multiple angles by using the Radon inverse transform algorithm.

[0099] The free space sound field measurement method provided by the embodiment of the application specifically includes the following steps: Step 1: Experimental optical path building.

[0100] Referring to Figure 3As shown, a set of free-path acoustic field measurement system based on the principle of Michelson interferometer is constructed. The system uses a 532 nm continuous laser with model MSI-FN-532-300mw as the light source. The laser beam first passes through the variable magnification laser beam expander with model VBE-532-2-10X for beam expansion to obtain a suitable beam diameter for observing the interference fringes. The expanded laser beam is incident on the non-polarizing beam splitter (BS) with model BS2155-A. The beam splitter divides the laser beam into two paths: one as the measurement light and the other as the reference light. The measurement light passes through the region where the acoustic field to be measured is placed and is incident on the high-power laser line mirror M1 with model LPM10-532-1064-HP and is reflected back to the beam splitter. The reference light is incident on another mirror M2 with model LPM10-532-1064-HP and is reflected back to the beam splitter. The two reflected lights recombine at the beam splitter, interfere with equal thickness, and form interference fringes with alternating bright and dark. The interference fringes are enlarged by the objective lens and recorded by the scientific-grade CCD camera. All optical elements are accurately fixed and adjusted by the high-stability series adjustment stand.

[0101] Step 2: Acoustic field measurement data acquisition.

[0102] A Matlab control framework is built in the computer to realize the control of the scanning path. The loudspeaker is tightly fixed on the stepper motor, which drives it to move and rotate.

[0103] 1. Acquisition of reference image: In the absence of acoustic field, an interference fringe image is acquired by the CCD camera as the reference image.

[0104] 2. Introducing acoustic field and scanning: The acoustic field is generated by the loudspeaker and placed in the measurement path of the Michelson interferometer, so that the measurement light passes through the acoustic field. As shown in Figure 4 , the scanning process includes: Translation scanning: As shown in Figure 4 (a), the laser beam position is kept unchanged, and the loudspeaker is stepwise translated in the two-dimensional plane. After each translation by a certain distance, the CCD camera acquires the interference fringe pattern at the current position (see Figure 5 ).

[0105] Rotation scanning: As shown in Figure 4 (b), based on the translation scanning or alone, the rotation scanning is performed to disturb the laser beam at different angles. After each rotation by a certain angle, the CCD camera acquires the corresponding interference fringe pattern (see Figure 5 ).

[0106] 3. Repeat the above translation-scanning and rotation-scanning processes to complete the data acquisition of "acoustic field interference fringe patterns" at multiple positions and angles.

[0107] Step 3: Interference fringe image analysis and light phase change extraction.

[0108] For each pair of "sound field-free interference fringe pattern" and "sound field interference fringe pattern" collected in step 2, the following processing is performed: 1. Image preprocessing: Perform filtering, denoising and other preprocessing operations on the collected interference pattern to improve image quality.

[0109] 2. Fringe edge detection: Use edge detection techniques in image processing to identify the clear edges of the interference fringes. This algorithm locates the edge information in the image by detecting the abrupt change in pixel gray value (such as object contour, texture change, etc.). In order to improve the measurement accuracy, a sub-pixel edge detection algorithm can be used to analyze the fringe displacement and accurately obtain the actual movement distance of the interference fringes. The figure shows the actual observed interference fringes.

[0110] 3. Light phase change calculation: According to the linear relationship between the movement of interference fringes and the optical path difference, and the relationship between the change of optical path difference and the light phase change, the fringe movement is converted into the light phase change caused by the sound field when the light beam passes through the sound field.

[0111] The edge detection algorithm locates the edge information in the image by detecting the abrupt change in pixel gray value, including object contour and texture change. The sub-pixel edge detection algorithm is used to analyze the fringe displacement and obtain the actual movement distance of the interference fringes. The four-stage edge detection algorithm of Gaussian filtering, Sobel operator gradient calculation, non-maximum suppression and double threshold segmentation is suitable for the sub-pixel level displacement measurement requirement of interference fringes, and the mathematical principles of each stage are as follows: For the random noise (such as light intensity fluctuation, electronic noise) in the interference fringe image collected by CCD, a two-dimensional Gaussian function is used to smooth the image as follows: ; ; In the formula, is a two-dimensional Gaussian function, which is a kernel function for smoothing the image, is an exponential function with a natural constant e as the base, which is the core component of the two-dimensional Gaussian function, used to describe the decay characteristics of the function value with x , y . is the coordinate variable of the image in the horizontal direction, used to determine the position-related value of the two-dimensional Gaussian function in the horizontal direction, is the coordinate variable of the image in the vertical direction, used to determine the position-related value of the two-dimensional Gaussian function in the vertical direction, is the standard deviation of the Gaussian kernel, for balancing noise suppression and edge blurring; is the filtered image gray value, is the original interference fringe image gray value, is the convolution operation; the gradient of the filtered image in x and y directions is calculated by Sobel operator to determine the amplitude and direction of the edge as follows: wherein, are the gradient in x and y directions respectively, is the gradient amplitude, which is used to reflect the edge intensity; is the gradient direction, which is used to reflect the edge orientation and match the vertical direction of the interference fringe; the local extremum of the gradient amplitude is judged along the gradient direction to retain the edge center pixel and eliminate the non-edge pixel as follows: wherein, is the non-maximum suppression image, the edge width is refined from 3 to 5 pixels to 1 pixel, the broken edges are connected by setting high and low threshold values, is the two-dimensional Gaussian function, is the adjacent pixel in the gradient direction; finally, the sub-pixel edge positioning is used to improve the accuracy, the sub-pixel level edge coordinate calculation is realized based on the interpolation method, the gray values of 3x3 neighborhood around the edge are fitted by using the quadratic polynomial, the gray values around the edge satisfy , the derivative of is calculated, and the derivative is 0, the sub-pixel coordinates are obtained as follows: wherein, is the polynomial fitting coefficient, is the sub-pixel level edge coordinate, the positioning accuracy is 0.1 pixel; is the pixel coordinate.

[0112] ​​​​​​​​The edge detection algorithm of the application is improved through Gaussian filter optimization and sub-pixel positioning, solving the problems of edge blur, low positioning accuracy and poor efficiency in traditional methods in interference fringe measurement. The edge positioning error of the traditional method is about 1.2 pixels, and the edge positioning error of the improved algorithm is about 0.1 pixels, which is 10 times higher than that of the traditional detection method, and also reduces the measurement error of the fringe movement. The traditional edge detection algorithm is a general type (such as suitable for natural images), and the algorithm of the application combines the periodicity and vertical trend characteristics of the interference fringe, limits the gradient direction calculation to 0° or 90° (only vertical edge is needed), and reduces the operation amount by 50%; through the combination of sub-pixel positioning and fast fitting, the processing efficiency is improved by 2 times while the accuracy is improved by 10 times. The algorithm of the application can directly process low-contrast fringe images collected by a CCD, while the traditional method needs to additionally enhance the image contrast, which is more suitable for actual sound field measurement scenes.

[0113] 1) Image acquisition and region cropping: interference fringe images are acquired from a CCD camera, and the image is selectively cropped by a region of interest to focus on the core area of the interference fringe; Supplement: The interference fringe directly acquired by the CCD camera is the image shot by the complete CCD target surface, which includes unnecessary black edges and even interference of blemishes. The interference fringe blur phenomenon at the edge of the picture interferes with edge extraction, so only the region with clear and good quality interference fringe in the picture needs to be cropped. The code of the frame selection algorithm is written by Matlab software, and after the cropped picture area is determined, the program crops and saves the region with good interference quality in each image.

[0114] 2) Image filtering and gradient calculation: apply an edge detection filter to the cropped image to obtain the edge gradient and direction of the image; 3) Edge extraction and threshold processing: after applying the filter, the high-light edge in the fringe image is extracted through threshold processing of the threshold operator or the hysteresis_threshold operator to form the edge fine line; 4) Non-maximum suppression and edge thinning: the extracted edge is subjected to non-maximum suppression, and the skelcton operator is used to draw or thin the edge; the discontinuous edge is subjected to merging and background separation processing to obtain a complete contour; 5) Sub-pixel edge positioning: a sub-pixel edge detection algorithm is used to realize the sub-pixel level accurate positioning of the interference fringe edge by interpolating or curve fitting the gray information between pixels; 6) Fringe displacement calculation: Obtain the sub-pixel level edge position of the reference image in the sound field, and the sub-pixel level edge position of each scanning position and angle image in the sound field; By comparing the two sets of edge data, the actual displacement distance of each interference fringe in the horizontal or vertical direction is calculated.

[0115] Step 4: Sound field sound pressure reconstruction.

[0116] The light phase change data obtained in step 3 (i.e. the path integral information of the sound field) at different positions and angles is input into the sound field reconstruction algorithm. The present application adopts Radon inverse transform algorithm for sound field reconstruction. Radon inverse transform can inverse the original two-dimensional or three-dimensional sound field spatial distribution from multi-angle line integral projection data. Through this algorithm, the sound field integral information carried by the laser beam can be converted into the sound pressure value of each point in the real space, so as to reconstruct the complete free field sound pressure distribution map.

[0117] First, the reconstruction performance of the present method at audible frequencies is simulated in a 2kH scenario. Due to the relatively low frequency, the sound wave wavelength is relatively long, and the sound pressure distribution presents a wide concentric circular ripple. The reconstructed image clearly captures the concentric circular structure of the original sound field, with high image clarity and high consistency with the original sound field. The final calculated root mean square error (RMSE) is 1.0124. This low error value indicates that the present method has high precision and reliability in reconstructing audible sound fields.

[0118] A complex sound field generated by the superposition of two independent sound sources is simulated to verify the processing capability of the present method for multi-source sound fields. Sound source 1 is 1000 Hz, located at (-0.2m, -0.2m), and sound source 2 is 2000 Hz, located at (0.2m, 0.2m). The vibration velocity amplitude is 20 m / s. As shown in the figure, the original sound field presents two independent sound pressure centers, and the two sets of concentric circular ripples are superimposed to form a typical complex interference pattern. The reconstructed image clearly identifies and reconstructs the relative position and intensity of the two sound sources, effectively capturing the complex sound field superposition pattern. The final calculated root mean square error (RMSE) is 2.0350. This low error value indicates that the present method can successfully process complex sound fields composed of multiple sound sources and accurately reconstruct them.

[0119] In the ultrasonic field (200 kHz) scene, the reconstruction performance of the method of the present application is simulated at very high frequencies. Sound wave frequency: 200 kHz, sound source vibration amplitude: 20 m / s, sound source z coordinate: -0.1 m. As shown in the figure, due to the extremely high frequency, the sound wave wavelength is very short, resulting in extremely dense and complex ring-shaped ripples in the spatial distribution of sound pressure. Although the reconstructed image captures the basic ring structure, due to the excessive density of the ripples, the reconstructed details are lost, and the image is relatively blurred. The final calculated root mean square error (RMSE) is 101.2353. This high error value indicates that without increasing the scanning resolution (i.e., reducing the translation step and angular interval), the reconstruction accuracy of the method for extremely high frequency sound fields with extremely short wavelengths will be limited.

[0120] The method can effectively and accurately reconstruct low and medium frequency sound fields, including some complex sound fields. Although for extremely high frequency ultrasonic fields (200 kHz) with extremely short wavelengths, the reconstruction accuracy will decrease at the current resolution, but by increasing the sampling density (reducing the translation step and rotation angle), the reconstruction effect can be significantly improved. For complex scenes such as multi-point sound sources, the method shows strong processing capability and reconstruction accuracy.

[0121] Example inversion effect: 1. Acousto-optic effect measurement simulation: As Figure 6 shown, the change of interference fringes of the Michelson interferometer when receiving sound field disturbance was simulated. The simulation object is a point sound source plane sound field with a frequency of 1000 Hz and a diameter of 1 m, and the point sound source is located 10 cm below the measurement plane, with a light wavelength of 532 nm. The simulation results show that the sound pressure distribution leads to the change of air refractive index, which in turn causes the phase change of light transversely passing through the sound field. When the maximum sound pressure is 1.539 Pa, the maximum refractive index change is 2.96 x 10 -9 , and the maximum phase change is 0.027 rad.

[0122] The interference fringes formed by the interference of the measured light and the reference light after passing through the sound field were also simulated. At an equal thickness interference mirror angle of about 0.1 degrees, the optical path difference changes in the range of 1.03-2.74 nm, the fringe movement amount ranges from 0.0032 to 0.0087 fringes, the fringe spacing is 0.16 mm, and the fringe displacement amount ranges from 0.51 to 1.37 μm. This shows that the fringe movement amount is in microns, which cannot be observed by the human eye and needs to be measured by a high-resolution CCD camera and edge detection algorithm.

[0123] Theoretical calculation shows that the measurement system must reach the accuracy of one thousandth of the wavelength. Under the condition of equal-thickness interference angle of 0.1° and laser wavelength of 532 nm, the equal-thickness interference fringe spacing is about 152 μm. The pixel size of the existing industrial camera in the laboratory is about 1.85 μm, and after magnification of 10 times by the objective lens, it is 0.18 μm / pixel. When the moving arm moves half the wavelength displacement, the fringe moves one spacing, and the displacement sensitivity is about 0.31 nm / pixel. Combined with optical magnification and sub-pixel edge detection algorithm, sub-nanometer level displacement resolution can be realized in theory. The light is affected by the sound field with a sound pressure of about 2 Pa, and the optical path changes in the nanometer level, so high-precision measurement of the sound field can be realized.

[0124] 2. Radon transform simulation verification: The principle of Radon transform is shown in Figure 7 . Computer simulation analyzes the acousto-optic effect as an explanation of the Radon transform of the sound field. From the simulation results, it can be seen that the sound pressure distribution image of the point sound source is similar to a concentric circle, which spreads from inside to outside and the sound pressure gradually attenuates. The projection distribution obtained after Radon transform looks like a fence with thick middle and thin sides. This is the result of the spherical symmetry of the sound field, that is, the integral of the sound field is independent of the angle, resulting in the same result. This verifies that the laser measurement of the sound field is actually a Radon transform process. Radon transform of the simply supported aluminum plate is simulated, further verifying the feasibility of reconstructing the sound field by laser measurement data.

[0125] These results show that the free-field sound field measurement system and method based on acousto-optic effect proposed by the present application have the feasibility of realizing high-precision and non-destructive sound field measurement in theory and simulation.

[0126] The above describes only the preferred specific embodiments of the present application, but the protection scope of the present application is not limited thereto, and any modification, equivalent replacement and improvement made by those skilled in the art within the technical scope disclosed by the present application shall be covered within the protection scope of the present application.

Claims

1. A method of free-field sound field measurement, characterized by, The method comprises the following steps: S1, light path building: constructing a free optical path sound field measurement system based on the principle of Michelson interferometer; S2, sound field data acquisition: collecting reference images, introducing a sound field to be measured, and performing translation and rotation scanning through a computer to control a loudspeaker or a laser beam to complete interference pattern data acquisition; S3, interference fringe image processing and light phase information extraction: preprocessing the collected interference pattern, identifying the edges of the interference fringes by using edge detection technology, and calculating the light phase change; S4, sound field sound pressure reconstruction: using Radon inverse transform algorithm for sound field reconstruction, converting the sound field integral information carried by the laser beam into the sound pressure value of each point in the real space, and reconstructing the free field sound pressure distribution map.

2. The method of claim 1, wherein, In step S1, the light path building comprises: The free optical path sound field measurement system uses a continuous laser as a light source, and the laser beam emitted by the continuous laser is expanded by a beam expander; The expanded laser beam is incident on a beam splitter and is divided into measurement light and reference light; The measurement light passes through the region where the sound field to be measured is placed, is incident on a mirror M1, and is reflected back to the beam splitter; the reference light is incident on a mirror M2 and is reflected back to the beam splitter; 3. The method of claim 2, wherein, The two beams of reflected light recombine at the beam splitter, undergo equal-thickness interference, and form interference fringes with light and dark alternating; the interference fringes are enlarged by an objective lens and recorded by a CCD camera. The sound field to be measured is generated by a loudspeaker, and all optical elements are accurately fixed and adjusted by a high-stability series adjustment stand; the laser, beam splitter, beam expander, mirror, and CCD are installed on the high-stability series adjustment stand and rigidly fixed on an optical vibration isolation platform; the adjustment stand is provided with a fine adjustment screw mechanism for adjusting the pitch angle, horizontal angle, and height to ensure that the light beam propagates along the central axis of the platform; after fixing, the adjustment stand locking screw is tightened to prevent the elements from being angularly offset due to slight vibration; after the laser is fixed, the pitch angle and horizontal position are adjusted to make the light beam propagate along the center of the optical platform; after the beam expander is installed, the height is accurately adjusted by the adjustment stand to make the expanded light spot uniformly cover the center of the beam splitter; The beam splitter is fixed on the adjustment stand, and the position is controlled to be about 10-15 cm behind the beam expander, and the height and angle are adjusted to make the reflected light and transmitted light vertically incident on the centers of the mirror M1 and the mirror M2, respectively; 4. The method of claim 1, wherein, The heights and angles of the mirrors M1 and M2 are adjusted to make the reflected light coincide with the incident light when returning, and clear interference fringes are formed at the beam splitter. In step S2, the sound field data acquisition comprises: Collecting reference images: in the absence of sound field interference, the initial interference fringe pattern is recorded by a CCD camera as a reference image for subsequent data processing; Introducing a sound field: introducing the sound field to be measured into the measurement light path so that the sound field disturbance passes through the measurement light in the region; Scanning measurement: translation and rotation scanning are controlled by a computer; The translation-scanning and rotation-scanning processes are repeated to complete interference pattern data acquisition at multiple positions and angles.

5. The method of claim 4, wherein, The scanning measurement includes: fixing the laser beam position, stepwise translation and rotation of the loudspeaker in a two-dimensional plane; the loudspeaker is fixed on a stepping motor, and the loudspeaker is moved by the motor; after each translation by a certain distance or rotation by a certain angle, the interference fringe pattern after the sound field disturbance is photographed and recorded in real time by the CCD camera.

6. The method of claim 1, wherein, In step S3, the interference fringe image processing and light phase information extraction include: (1) image preprocessing: filtering and denoising preprocessing operation is performed on the collected interference pattern; (2) fringe edge detection: an edge detection algorithm is used to identify the edges of the interference fringes; (2) light phase change calculation: according to the linear relationship between the movement of the interference fringes and the optical path difference, and the relationship between the change of the optical path difference and the light phase change, the fringe movement is converted into the light phase change caused by the sound field.

7. The method of claim 6, wherein, The edge detection algorithm locates the edge information in the image by detecting the mutation area of the pixel gray value, and the mutation area includes the object contour and texture change; the sub-pixel edge detection algorithm is used to analyze the fringe displacement, and the actual movement distance of the interference fringes is obtained; the four-stage edge detection algorithm of Gaussian filtering, Sobel operator gradient calculation, non-maximum suppression and double threshold segmentation is suitable for the sub-pixel level displacement measurement requirement of the interference fringes, and the specific process is as follows: For the random noise of the interference fringe image collected by the CCD, including light intensity fluctuation and electronic noise, a two-dimensional Gaussian function is used to smooth the image, and the expression is: ； ； wherein is a two-dimensional Gaussian function, which is a kernel function for smoothing the image, is an exponential function with base e , which is a core component of the two-dimensional Gaussian function, used to describe the decay property of the function value with respect to x , y ; is a coordinate variable of the image in the horizontal direction, used to determine the position-dependent value of the two-dimensional Gaussian function in the horizontal direction, is a coordinate variable of the image in the vertical direction, used to determine the position-dependent value of the two-dimensional Gaussian function in the vertical direction, is the standard deviation of the Gaussian kernel, , used to balance the noise suppression and edge blurring; is the gray value of the filtered image, is the gray value of the original interference fringe image, is the convolution operation; The gradient of the filtered image in the direction of Sobel operator is calculated The amplitude and direction of the edge are determined, and the expression is as follows: ； ； ； ； In the formula, respectively direction gradient, is a gradient amplitude, used to reflect the edge intensity; is a gradient direction, used to reflect the edge direction and match the vertical direction of the interference fringes. Local extremum judgment is performed on the gradient amplitude along the gradient direction, the edge center pixel is retained, and the non-edge pixel is removed, and the expression is: ； wherein is the non-maxima suppression image, the edge width is refined from 3 to 5 pixels to 1 pixel, the broken edges are connected by setting a high threshold and a low threshold, is the adjacent pixel in the gradient direction; The sub-pixel edge positioning is used to improve the precision, the interpolation method is used to realize the sub-pixel level edge coordinate calculation, a quadratic polynomial is used to fit the 3*3 neighborhood gray value near the edge, the gray value near the edge is set to satisfy , the derivative is obtained, and the derivative is 0, so that the sub-pixel coordinates are obtained. ​ ； In the formula, is a polynomial fitting coefficient, is a sub-pixel level edge coordinate, and the positioning accuracy is 0.1 pixels; is a pixel coordinate.

8. The method of claim 7, wherein, The actual displacement distance of each interference fringe in the horizontal or vertical direction includes: The interference fringe spacing of the Michelson interferometer in the equal thickness interference case is: ； wherein is the distance between the interference fringes, is the wavelength of the laser light, is the mirror angle; Stripe movement amount Is: ； wherein is the optical path difference changed when the light path passes through the acoustic field; Actual movement distance of the stripe Is: ； In the formula, is the phase variation amount; By the phase change amount A linear relationship between the actual distance moved by the fringe and the distance moved by the fringe, enabling the sound field to be back-calculated.

9. The method of claim 1, wherein, In step S4, the sound field sound pressure reconstruction includes: inputting the obtained light phase change data at different projection angles and positions into the Radon inverse transform algorithm for sound field reconstruction to generate a free field sound pressure distribution map; The reference frame of the sound field is The coordinate system, rotates The coordinate system after the rotation of the angle becomes The coordinate system, The coordinate system and The transformation relationship between the coordinate systems is: ； The projection integral process is: ； wherein is the sound pressure to be measured in the sound field, is the sound pressure along the direction path is the integral of the sound pressure, is the path of the light through the sound pressure perturbation region; According to the dynamic offset, the relationship between the Radon transform of the sound pressure and the sound pressure in the time domain is: ； wherein is the wave number of the light propagating in vacuum, is the pressure birefringence of the medium, is the phase measured by the interferometer, is the isentropic pressure birefringence, is the integral of the acoustic pressure in the direction along the path , is the reconstructed acoustic pressure at the coordinates and time . ； wherein is a relative refractive index parameter of the medium, is a coefficient related to the properties of the medium, is a reference pressure; The measured sound field sound pressure is reconstructed by Radon inverse transform as: ； where is the Radon inverse transform operator, which acts to backpropagate the projection integral results to the sound pressure distribution.

10. A free-field sound field measurement system characterized by, The system is used for regulating and controlling the free space sound field measurement method of any one of claims 1-9, and the system comprises: A laser is used to emit a laser beam with good coherence, preferably a continuous laser; A beam expander is used to expand the laser beam; A beam splitter is used to divide the expanded laser beam into measurement light and reference light; A mirror is used to reflect the measurement light and the reference light to cause re-converging interference; wherein the mirror M1 is placed at the end of the sound field to be measured, so that the measurement light passes through the sound field and is reflected; A loudspeaker / sound source is placed in the measurement light path as the sound field to be measured, for generating sound wave disturbance; A scientific grade CCD camera is placed in the interference fringe forming area for recording the interference pattern; An objective lens is used to magnify the interference fringes and improve the image resolution recorded by the CCD camera; Computer and control system for controlling the scanning motion of the loudspeaker / sound source, the data acquisition of the CCD camera and performing the image processing and sound field reconstruction algorithm. High stability series of adjustment mounts for precise fixing and adjustment of optical element positions.

Images