Single-camera three-dimensional particle tracking speed measurement method and system based on double-helix point spread function

Abstract

The invention discloses a single-camera three-dimensional particle tracking velocity measurement method and system based on a double-helix point spread function. The test method comprises the following steps: exciting tracer particles in a flow field to be measured; performing double-helix phase modulation on the particle scattered light; calibrating the angle-depth relation of the double-helix particles; collecting an original image sequence of flow field particles; performing three-dimensional particle field reconstruction on the original image sequence; carrying out three-dimensional particle velocity field solving on the three-dimensional particle field; and outputting a high-precision three-dimensional velocity field. According to the test method, double-helix phase modulation is carried out on particle scattered light, the depth of particles is determined through angle information of two modulated lobes, particle three-dimensional positioning of a single camera is achieved, and then three-dimensional flow field particle tracking speed measurement of the single camera is achieved. Compared with a single-light-field camera three-dimensional particle tracking velocity measurement technology, the spatial resolution is improved. Compared with a three-hole mask three-dimensional particle tracking speed measurement technology, the light utilization rate is improved. The method is suitable for high-precision speed measurement of a three-dimensional flow field.

Description

Technical Field

[0001] The present invention relates to the field of optical measurement methods, and in particular to a single-camera three-dimensional particle tracking velocity measurement method and system based on a double-helix point spread function. Background Art

[0002] Particle tracking velocimetry (PTV) is an image-based flow measurement technology that can collect and reproduce flow characteristics in a flow field by spreading tracer particles in the flow field, using an imaging system to collect particle images, and using image calculations to track and measure particles. For two-dimensional flow measurement, only a single traditional camera plus reasonable matching of imaging parameters such as aperture, exposure time, and focal length are needed to better realize the collection and reconstruction of a two-dimensional flow field.

[0003] As for three-dimensional flow measurement, it is often difficult to achieve depth resolution of particles within the field of view due to the limitations of traditional camera imaging principles. Multi-camera particle tracking velocimetry methods such as tomographic PTV and stereo PTV use multiple cameras to synchronously collect the target flow field and use the angle information obtained when multiple cameras image the same target particle to achieve three-dimensional reconstruction of the target particle. Further, combined with the three-dimensional particle tracking algorithm, the velocity field reconstruction of the three-dimensional flow field is achieved. However, the three-dimensional particle tracking velocimetry method driven by multiple cameras has high requirements for transmission bandwidth and camera synchronization, and its multi-camera system calibration is often more complicated. In order to reduce system cost and system complexity, single-camera three-dimensional imaging technologies such as light field PTV and three-hole mask PTV have been continuously developed in recent years. Light field PTV places a microlens array behind the imaging lens and uses a single camera to obtain multi-angle information, thereby realizing single-camera flow measurement of three-dimensional flow fields. However, the introduction of the microlens array greatly reduces the spatial resolution of the imaging system, and light field imaging also has the problem of low axial accuracy caused by the axial stretching of the reconstructed particles. The three-hole mask PTV places a three-hole mask in front of the imaging lens, and each hole position realizes particle imaging at an angle. Using the angle information generated by the three-hole mask, the three-hole mask PTV can also realize three-dimensional measurement of the flow field. However, due to the introduction of the three-hole mask, the light utilization efficiency of the imaging system is greatly reduced, which puts forward greater requirements on the power of the laser light source in practical applications. Summary of the invention

[0004] Purpose of the invention: In view of the shortcomings of existing single-camera three-dimensional imaging technology, such as low spatial resolution and poor light utilization, the present invention provides a single-camera three-dimensional particle tracking velocity measurement method and system based on a double-helix point spread function.

[0005] Technical solution: To solve the above problems, the present invention adopts a single-camera three-dimensional particle tracking velocity measurement method based on a double-helix point spread function, which includes the following steps:

[0006] Step 1: collimate the laser light source and excite the tracer particles in the flow field to be measured;

[0007] Step 2: using an imaging system to collect particle scattered light, and using a phase-type spatial light modulator to perform double-helix phase modulation on the particle scattered light;

[0008] Step 3, calibrating the angle-depth relationship of the two lobes of a single double helix to obtain an angle-depth calibration curve;

[0009] Step 4: Using an imaging system to collect a time image sequence of flow field particles to obtain a time image sequence of original double helical particles;

[0010] Step 5, reconstructing the original double-helix particle time image sequence into three-dimensional particles to obtain a three-dimensional particle field time image sequence;

[0011] Step 6: Use the Lagrangian particle tracking algorithm to solve the three-dimensional particle field time image sequence to obtain the three-dimensional particle velocity field;

[0012] Step 7: Output high-precision three-dimensional particle velocity field.

[0013] Furthermore, the double-helix phase modulation of the particle scattered light using a phase-type spatial light modulator as described in step 2 is specifically performed as follows: the phase-type spatial light modulator is placed at the Fourier plane of the 4f lens group and a double-helix phase image is loaded to phase modulate the particle scattered light excited by the laser.

[0014] Furthermore, the specific operation of step three is: placing a fixed particle in the focal plane of the imaging system, the particle is located at the center of the field of view of the imaging system, moving the fixed particle along the main optical axis of the imaging system, collecting the particle calibration images after phase modulation at different depth positions, and establishing the calibration relationship curve f between the depth and the angle of the double helix particle z (θ).

[0015] Furthermore, the specific operation of step five is: perform double Gaussian fitting on the original double helix particle image, and use the fitting results to determine the center coordinates and angles of the two lobes of the double helix; use the center coordinates of the two lobes of the double helix to determine the xy coordinates of the three-dimensional particle, and use the angle information of the two lobes of the double helix combined with the angle-depth calibration curve to determine the z coordinate of the three-dimensional particle.

[0016] Furthermore, the double Gaussian fitting is performed on the original double helix particle image, and the specific operation is: approximating the intensity distribution of the double helix particles to the following double Gaussian distribution function:

[0017]

[0018] Among them, (u, v) is the image pixel coordinate, A1 is the amplitude of the first lobe of the double helix, (μx1 , μ y1 ) is the position of the first lobe, A2 is the amplitude of the second lobe of the double helix, (μ x2 , μ y2 ) is the position of the second lobe, σ x1 , σ y1 is the transverse and longitudinal width of the first lobe, σ x2 , σ y2 is the horizontal width and vertical width of the second lobe, and B is the background image;

[0019] Replace A1, A2, (μ x1 , μ y1 )、(μ x2 , μ y2 ),σ x1 , σ y1 , σ x2 , σ y2 , B are taken as the unknown parameters, and the original double helix particle image intensity distribution is used as the initial value. The unknown parameters are determined by least square fitting to obtain the above double Gaussian distribution function.

[0020] Furthermore, the fitting results are used to determine the center coordinates (μ x , μ y ) and the angle θ are calculated as follows:

[0021]

[0022] The xy coordinates of the three-dimensional particle are determined by using the center coordinates of the two lobes of the double helix, and the z coordinate of the three-dimensional particle is determined by using the angle information of the two lobes of the double helix combined with the angle-depth calibration curve. The specific calculation is as follows:

[0023]

[0024] Complete the three-dimensional particle reconstruction and obtain the three-dimensional particle field time image sequence.

[0025] Furthermore, the specific operations of step six are as follows: first, solve the minimum total standard deviation of the particle trajectory in the time image sequence and perform particle matching of multiple frames; second, fit the matched particle time series into the particle motion trajectory; finally, use the inter-frame time and particle trajectory coordinates to calculate the velocity field of the flow field.

[0026] Furthermore, the method of solving the minimum total standard deviation of particle trajectories in the time image sequence and performing particle matching of multiple frames is as follows: first, extract the three-dimensional coordinates of the particles in the first four frames, and solve the length standard deviation σ between the particle in the first frame and the neighboring particles at the coordinates of the particle in the second to fourth frames. length and the angle standard deviation σθ , and then further solve the total standard deviation of particle matching σ all :

[0027]

[0028] σ length is the standard deviation of the mean vector length, σ θ is the standard deviation of the mean vector angle, l ab is the length of the displacement vector formed by the first frame particles and the second frame particles, l bc is the length of the displacement vector formed by the particles in the second frame and the particles in the third frame, l cd is the length of the displacement vector formed by the particles in the third frame and the particles in the fourth frame, is the average displacement distance of the particles, θ ac for l ab With l bc The angle, θ bd for l bc With l cd The angle of is the average angle, σ all is the total standard deviation; the total standard deviation σ all The smallest four-frame particle trajectory is used as the optimal particle trajectory to complete the four-frame particle matching. This method is repeated to complete the particle matching of all time series particle images.

[0029] Furthermore, the method of using the inter-frame time and particle trajectory coordinates to calculate the velocity field of the flow field specifically comprises the following operations: using the particle trajectory coordinates and the inter-frame time obtained by particle matching, solving for the average three-dimensional velocity vector of the particle within the inter-frame time, and combining the particle trajectory coordinates to obtain the velocity field of the flow field.

[0030] The present invention also provides a test system for the three-dimensional particle tracking velocity measurement method, comprising:

[0031] (1) A single-camera three-dimensional particle tracking imaging system, including an industrial camera, a phase-type spatial light modulator, an imaging lens, a non-polarizing beam splitter cube, a linear polarizer, an aperture, and a 4f system lens group;

[0032] (2) Particle excitation system, including laser, collimating and beam expanding lens group;

[0033] (3) A data processor is used to reconstruct a three-dimensional velocity field of the flow field by obtaining a sequence of particle images in the flow field.

[0034] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages: (1) without sacrificing spatial resolution, it can realize three-dimensional flow field measurement with a single camera, thus solving the problem of low spatial resolution of light field PTV; (2) without introducing a three-hole mask and thus greatly reducing the light utilization rate, it can realize three-dimensional flow field measurement with a single camera, thus solving the problem of low light utilization rate of three-hole mask PTV; (3) a single camera can replace multiple camera systems or special optical elements, thus reducing the cost of the three-dimensional PTV system, greatly simplifying the configuration and operation of the system, and effectively promoting the high-precision measurement research of three-dimensional flow fields. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is a schematic diagram of the overall process of the speed measurement method of the present invention;

[0036] Figure 2 Schematic diagram of the overall structure of the speed measurement system of the present invention;

[0037] Figure 3 It is a schematic diagram of simulation of double helix point spread function of phase diagrams with different topological charges in the spatial light modulator of the present invention;

[0038] Figure 4 The calibration relationship curve f of the depth and angle of the double helix particle of the present invention is z (θ);

[0039] Figure 5 Schematic diagram of the principle of the speed measurement method of the present invention;

[0040] In the figure: 1: laser, 2: collimating and expanding lens group, 3: single camera three-dimensional particle tracking imaging system, 4: data processor, 5: imaging lens, 6: linear polarizer, 7: 4f system 60mm lens, 8: non-polarizing beam splitter cube, 9: aperture, 10: phase-type spatial light modulator, 11: 4f system 60mm lens, 12: industrial camera. DETAILED DESCRIPTION

[0041] like Figure 1 and Figure 2 As shown, a single-camera three-dimensional particle tracking velocity measurement method based on a double-helix point spread function in this embodiment includes the following steps:

[0042] Step 1: Collimate the laser light source and excite the tracer particles in the flow field to be measured.

[0043] Step 2: Use the imaging system to collect the particle scattered light, and use the phase-type spatial light modulator to perform double-helix phase modulation on the particle scattered light. The specific structure of the imaging system is as follows: Figure 2As shown in (b), the 4f lens group includes two 60mm lenses. The phase-type spatial light modulator is placed at the Fourier plane of the 4f lens group and loaded with a double-helix phase pattern to phase modulate the particle scattered light excited by the laser.

[0044] Step 3: Calibrate the angle-depth relationship of the two lobes of a single double helix to obtain an angle-depth calibration curve. Place a fixed particle in the focal plane of the imaging system. The particle is located at the center of the field of view of the imaging system. Move the fixed particle along the main optical axis of the imaging system to collect phase-modulated particle calibration images at different depths, such as Figure 3 As shown, n = 0 is the imaging result of the original particle at different depths, and n = 15 is the particle imaging result after double helix phase modulation, where n is the topological charge of the double helix phase diagram, and the left side is the double helix phase diagram when the topological charge n = 15. A calibration relationship curve f between depth and angle of double helix particles is established z (θ), such as Figure 4 As shown, where θ is the angle of the double helix particle and z is the depth of the particle.

[0045] Step 4: Use the imaging system to collect the time image sequence of the flow field particles to obtain the time image sequence of the original double helix particles.

[0046] Step 5: Figure 5 As shown in the three-dimensional particle reconstruction steps in , the time image sequence of the original double helix particles is reconstructed into three-dimensional particles to obtain a three-dimensional particle field time image sequence.

[0047] First, a double Gaussian fitting is performed on the original double helix particle image. The intensity distribution of the original double helix particle image is used as the initial value and the intensity distribution of the double helix particles is approximated to the following double Gaussian distribution function through least squares fitting:

[0048]

[0049] Among them, (u, v) is the image pixel coordinate, A1 is the amplitude of the first lobe of the double helix, (μ x1 , μ y1 ) is the position of the first lobe, A2 is the amplitude of the second lobe of the double helix, (μ x2 , μ y2 ) is the position of the second lobe, σ x1 , σ y1 is the transverse and longitudinal width of the first lobe, σ x2 , σ y2 are the horizontal and vertical widths of the second lobe, and B is the background image.

[0050] Replace A1, A2, (μ x1 , μ y1 )、(μx2 , μ y2 ),σ x1 , σ y1 , σ x2 , σ y2 , B as the unknown parameters, use the original double helix particle image intensity distribution as the initial value, and determine the unknown parameters through least square fitting to obtain the above double Gaussian distribution function, and further obtain the double Gaussian particle distribution.

[0051] Secondly, the fitting results are used to determine the center coordinates of the two lobes of the double helix (μ x , μ y ) and angle θ:

[0052]

[0053] Finally, using the center coordinates of the two lobes of the double helix (μ x , μ y ), determine the xy coordinates of the three-dimensional particle, and use the angle θ of the two lobes of the double helix combined with the angle-depth calibration curve to determine the z coordinate of the three-dimensional particle:

[0054]

[0055] Complete the three-dimensional particle reconstruction and obtain the three-dimensional particle field time image sequence.

[0056] Step 6: Figure 5 As shown in the three-dimensional Lagrangian particle tracking steps in , the Lagrangian particle tracking algorithm is used to solve the three-dimensional particle field time image sequence to obtain the three-dimensional particle velocity field.

[0057] First, the minimum total standard deviation of particle trajectories in the time image sequence is solved to match particles in multiple frames. The three-dimensional coordinates of the particles in the first four frames are extracted, and the length standard deviation σ between the particle in the first frame and the neighboring particles at the coordinates of the particle in the second to fourth frames is solved. length and the angle standard deviation σ θ , and then further solve the total standard deviation of particle matching σ all :

[0058]

[0059] σ length is the standard deviation of the mean vector length, σ θ is the standard deviation of the mean vector angle, l ab is the length of the displacement vector formed by the first frame particles and the second frame particles, l bc is the length of the displacement vector formed by the particles in the second frame and the particles in the third frame, l cd is the length of the displacement vector formed by the particles in the third frame and the particles in the fourth frame, is the average displacement distance of the particles, θ ac for l ab With l bc The angle, θ bd for l bc With l cd The angle of is the average angle, σ all is the total standard deviation; the total standard deviation σ all The smallest four-frame particle trajectory is used as the optimal particle trajectory to complete the four-frame particle matching. This method is repeated to complete the particle matching of all time series particle images.

[0060] Secondly, the matched particle time series is fitted into the particle motion trajectory.

[0061] Finally, the particle trajectory coordinates and inter-frame time obtained by particle matching are used to solve the average three-dimensional velocity vector of the particle within the inter-frame time. Combined with the particle trajectory coordinates, the flow field velocity field is obtained.

[0062] Step 7: Output high-precision three-dimensional particle velocity field.

[0063] This embodiment also provides a test system for the above-mentioned single-camera three-dimensional particle tracking velocity measurement method, including:

[0064] (1) A single-camera three-dimensional particle tracking imaging system, including an industrial camera, a phase-type spatial light modulator, an imaging lens, a non-polarizing beam splitter cube, a linear polarizer, an aperture, and a 4f system lens group, wherein the 4f system lens group includes two 60 mm lenses.

[0065] (2) Particle excitation system, including laser and collimating and beam expanding lens group.

[0066] (3) A data processor is used to reconstruct a three-dimensional velocity field of the flow field by obtaining a sequence of particle images in the flow field.

Claims

1. A single-camera three-dimensional particle tracking velocity measurement method based on a double-helix point spread function, characterized in that: The following steps are involved: Step 1: collimate the laser light source and excite the tracer particles in the flow field to be measured; Step 2: using an imaging system to collect particle scattered light, and using a phase-type spatial light modulator to perform double-helix phase modulation on the particle scattered light; Step 3, calibrating the angle-depth relationship of the two lobes of a single double helix to obtain an angle-depth calibration curve; Step 4: Using an imaging system to collect a time image sequence of flow field particles to obtain a time image sequence of original double helical particles; Step 5, reconstructing the original double-helix particle time image sequence into three-dimensional particles to obtain a three-dimensional particle field time image sequence; Step 6: Use the Lagrangian particle tracking algorithm to solve the three-dimensional particle field time image sequence to obtain the three-dimensional particle velocity field; Step 7: Output high-precision three-dimensional particle velocity field.

2. The single-camera three-dimensional particle tracking velocimetry method according to claim 1, characterized in that: The double-helix phase modulation of the particle scattered light by using a phase-type spatial light modulator as described in step 2 is specifically performed as follows: the phase-type spatial light modulator is placed at the Fourier plane of the 4f lens group and a double-helix phase image is loaded to phase modulate the particle scattered light excited by the laser.

3. The single-camera three-dimensional particle tracking velocity measurement method according to claim 1, characterized in that: The specific operation of step three is as follows: place a fixed particle in the focal plane of the imaging system, the particle is located at the center of the field of view of the imaging system, move the fixed particle along the main optical axis of the imaging system, collect the particle calibration images after phase modulation at different depth positions, and establish the calibration relationship curve f between the depth and the angle of the double helix particle z (θ).

4. The single-camera three-dimensional particle tracking velocimetry method according to claim 1, characterized in that: The specific operations of step five are as follows: perform double Gaussian fitting on the original double helix particle image, and use the fitting results to determine the center coordinates and angles of the two lobes of the double helix; use the center coordinates of the two lobes of the double helix to determine the xy coordinates of the three-dimensional particle, and use the angle information of the two lobes of the double helix combined with the angle-depth calibration curve to determine the z coordinate of the three-dimensional particle.

5. The single-camera three-dimensional particle tracking velocity measurement method according to claim 4, characterized in that: The double Gaussian fitting is performed on the original double helix particle image. Specifically, the intensity distribution of the double helix particle is approximated to the following double Gaussian distribution function: Among them, (u, v) is the image pixel coordinate, A1 is the amplitude of the first lobe of the double helix, (μ x1 , μ y1 ) is the position of the first lobe, A2 is the amplitude of the second lobe of the double helix, (μ x2 , μ y2 ) is the position of the second lobe, σ x1 , σ y1 is the transverse and longitudinal width of the first lobe, σ x2 , σ y2 is the horizontal width and vertical width of the second lobe, and B is the background image; Replace A1, A2, (μ x1 , μ y1 )、(μ x2 , μ y2 ),σ x1 , σ y1 , σ x2 , σ y2 , B as the unknown parameters, use the original double helix particle image intensity distribution as the initial value, and determine the unknown parameters by least squares fitting to obtain the above double Gaussian distribution function; further, use the above double high-speed distribution function to determine the center coordinates (μ x , μ y ) and angle θ.

6. The single-camera three-dimensional particle tracking velocimetry method according to claim 5, characterized in that: The fitting results are used to determine the center coordinates (μ x , μ y ) and the angle θ are calculated as follows: The xy coordinates of the three-dimensional particle are determined by using the center coordinates of the two lobes of the double helix, and the z coordinate of the three-dimensional particle is determined by using the angle information of the two lobes of the double helix combined with the angle-depth calibration curve. The specific calculation is as follows: Complete the three-dimensional particle reconstruction and obtain the three-dimensional particle field time image sequence.

7. The single-camera three-dimensional particle tracking velocity measurement method according to claim 1, characterized in that: The specific operations of step six are as follows: first, solve the minimum total standard deviation of the particle trajectory in the time image sequence and perform particle matching of multiple frames; second, fit the matched particle time series into the particle motion trajectory; finally, use the inter-frame time and particle trajectory coordinates to calculate the velocity field of the flow field.

8. The single-camera three-dimensional particle tracking velocity measurement method according to claim 7, characterized in that: The method of solving the minimum total standard deviation of particle trajectories in the time image sequence and performing particle matching of multiple frames is as follows: first, extract the three-dimensional coordinates of the particles in the first four frames, and solve the length standard deviation σ between the particle in the first frame and the neighboring particles at the coordinates of the particle in the second to fourth frames. length and the angle standard deviation σ θ , and then further solve the total standard deviation of particle matching σ all : σ length is the standard deviation of the mean vector length, σ θ is the standard deviation of the mean vector angle, l ab is the length of the displacement vector formed by the first frame particles and the second frame particles, l bc is the length of the displacement vector formed by the particles in the second frame and the particles in the third frame, l cd is the length of the displacement vector formed by the particles in the third frame and the particles in the fourth frame, is the average displacement distance of the particles, θ ac for l ab With l bc The angle, θ bd for l bc With l cd The angle of is the average angle, σ all is the total standard deviation; the total standard deviation σ all The smallest four-frame particle trajectory is used as the optimal particle trajectory to complete the four-frame particle matching. This method is repeated to complete the particle matching of all time series particle images.

9. The single-camera three-dimensional particle tracking velocimetry method according to claim 6, characterized in that: The specific operation of calculating the velocity field of the flow field by using the inter-frame time and the particle trajectory coordinates is as follows: using the particle trajectory coordinates and the inter-frame time obtained by particle matching, solving for the average three-dimensional velocity vector of the particle within the inter-frame time, and combining the particle trajectory coordinates to obtain the velocity field of the flow field.

10. A test system for the three-dimensional particle tracking velocimetry method according to any one of claims 1 to 9, characterized in that: include: (1) A single-camera three-dimensional particle tracking imaging system, including an industrial camera, a phase-type spatial light modulator, an imaging lens, a non-polarizing beam splitter cube, a linear polarizer, an aperture, and a 4f system lens group; (2) Particle excitation system, including laser, collimating and beam expanding lens group; (3) A data processor is used to reconstruct a three-dimensional velocity field of the flow field by obtaining a sequence of particle images in the flow field.

Images