Background schlieren method flow field flow velocity measurement method based on pseudo-schlieren image

Abstract

The invention discloses a background schlieren method flow field flow velocity measurement method based on pseudo-schlieren images. The method comprises the following steps: constructing a background schlieren experiment measurement system; obtaining a speckle plate image sequence without a flow field background; obtaining a background distortion image sequence; obtaining a pseudo schlieren image sequence; and obtaining the final flow velocity of the flow field to be measured. The method has the advantages that a background plate distortion image sequence under the action of a flow field is collected through a high-speed camera, and pseudo-schlieren transformation processing is carried out through an improved mixed gradient operator and a multi-scale feature fusion technology; and a flow velocity field is solved in combination with a PIV-optical flow fusion algorithm, and real-time processing is realized through GPU parallel computing. The method breaks through the limitation that the traditional PIV technology depends on tracer particles, has the advantages of non-contact, full-field measurement, high temporal-spatial resolution and the like, and is particularly suitable for compressible flow field and turbulent flow field measurement. Experiments show that the method can realize high-resolution and high-frame-rate image processing, the uncertainty of flow velocity measurement is less than 3.2%, and a new technical means is provided for complex flow field diagnosis.

Description

Technical Field

[0001] The invention belongs to the technical field of aircraft engine tail jet flow field detection, and particularly relates to a background schlieren method flow field velocity measurement method based on pseudo schlieren images. Background Art

[0002] The flow characteristics of an aeroengine's exhaust flow field directly impact its performance, combustion efficiency, noise characteristics, and infrared radiation characteristics, making it a key research area in aeroengine design and optimization. Exhaust flow fields typically exhibit high temperatures, high velocities, and complex turbulent structures. Furthermore, multiple physical parameters, such as density, temperature, and velocity, are significantly coupled within the flow field, making measurement extremely challenging. Traditional contact measurement methods (such as thermocouples and pitot tubes) have significant limitations in high-temperature and high-velocity environments, are prone to interfering with the flow field itself, and struggle to provide high-resolution, full-field data.

[0003] In recent years, non-contact measurement techniques have attracted considerable attention due to their high applicability in complex flow environments. The background oriented schlieren (BOS) method, with its high precision, non-invasiveness, and rapid response, has become a crucial tool for studying high-temperature and high-speed flow fields. BOS technology, based on the optical deformation of light as it refracts through a density gradient, analyzes the offset of the background image to infer the density distribution in the flow field. By combining optical, thermodynamic, and fluid dynamics models, temperature and velocity information can be further derived.

[0004] However, in the near-field conditions of an aeroengine exhaust jet, light deflection angles are large, and the optical distortion of the background image is significant and nonlinear. Consequently, the accuracy of non-conventional BOS techniques in high-gradient flow fields is significantly reduced. Furthermore, the axisymmetric nature, complex turbulent structure, and multi-physics coupling of the exhaust jet flow field further complicate flow field measurement. Therefore, developing BOS techniques suitable for complex near-field flow fields and achieving velocity field reconstruction is a key research direction in aeroengine flow field measurement. Summary of the Invention

[0005] In order to solve the above problems, the object of the present invention is to provide a flow field velocity measurement method using a background schlieren method based on pseudo schlieren images.

[0006] In order to achieve the above object, the present invention provides a method for measuring flow field velocity using a background schlieren method based on a pseudo schlieren image, comprising the following steps performed in sequence:

[0007] 1) Constructing a background schlieren experimental measurement system, the system comprising: a high-speed camera, a flow field to be measured, a background speckle plate, and an optical adjustment platform; wherein the lens of the high-speed camera is directly facing the background speckle plate; the optical adjustment platform is disposed between the high-speed camera and the background speckle plate; the flow field to be measured is disposed on the optical adjustment platform and is located on the optical axis between the high-speed camera and the background speckle plate;

[0008] 2) Optically calibrate the background schlieren experimental measurement system. First, use a laser collimator and an optical adjustment platform to adjust the optical axis of the high-speed camera to be perpendicular to the background speckle plate. Then, use a checkerboard calibration plate to calibrate the internal parameters of the high-speed camera. Then, under conditions without flow field disturbance, use the high-speed camera to capture a reference image sequence of the background speckle plate. These reference image sequences are processed using PIVlab software to ensure that there is no obvious offset, thereby obtaining a flow-free background speckle plate image sequence.

[0009] 3) First, the flow field to be measured is started. After the flow field to be measured stabilizes, the high-speed camera is synchronously triggered to start acquiring images of the background speckle plate. During the acquisition process, the distorted image sequence of the background speckle plate under the action of the flow field is recorded in exposure mode. The signal-to-noise ratio and contrast of the distorted image are monitored in real time. Bad frames are detected immediately after the acquisition is completed. Finally, a high-temporal and spatial resolution background distorted image sequence that meets the requirements of subsequent processing is obtained and saved.

[0010] 4) Processing the background distortion image sequence obtained in step 3) and the flow field-free background speckle plate image sequence obtained in step 2) to obtain a two-dimensional pixel offset, then calculating the modulus of the speckle displacement field based on the two-dimensional pixel offset and performing normalized visualization processing on the images to obtain a visualized image sequence. Background noise reduction is then performed on the visualized image sequence using a non-local means filtering algorithm (NLM) to obtain a high-precision and high-resolution pseudo-schlieren image sequence;

[0011] 5) Based on the pseudo-schlieren image sequence obtained in step 4), combined with the real light path and the mapping relationship between the pixel offset on the camera imaging plane and the velocity vector in the flow field, the flow field velocity field is calculated using the improved PIV-optical flow algorithm to obtain the final flow velocity of the flow field to be measured.

[0012] In step 1), the high-speed camera has a frame rate of 8000 frames per second and is equipped with a telecentric lens with a focal length of 100 mm; the distance between the high-speed camera and the flow field to be measured is 1000 mm, and the distance between the flow field to be measured and the background speckle plate is 800 mm; the adjustment accuracy of the optical adjustment platform is ±0.1 mm.

[0013] In step 2), the background speckle plate has a dot density of 600 dots / mm 2 Random dot pattern;

[0014] The method for processing the reference image sequence using PIVlab software is as follows: first, the reference image sequence is subjected to Gaussian filtering for noise reduction and histogram equalization to enhance contrast, and then lens distortion correction is performed based on calibrated internal parameters, thereby completing the preprocessing of the reference image sequence; then, a multi-level iterative window cross-correlation algorithm is used, the calculation is accelerated by fast Fourier transform (FFT), a 50% overlap rate is set, and a Gaussian three-point fitting method is used to accurately detect correlation peaks; then, a normalized cross-correlation threshold NCC is used to filter out invalid vectors, and local median filtering is performed to eliminate outliers. Finally, the standard deviation σ of the full-field displacement vector is calculated to ensure a pass rate of more than 99.5%; wherein the normalized cross-correlation threshold NCC is greater than 0.8, and the standard deviation σ is less than 0.05 pixels.

[0015] In step 3), the exposure time of the high-speed camera is 100 μs, the background distortion image is saved in a RAW lossless format, and the resolution of each frame of the background distortion image is 1024×680 pixels.

[0016] In step 4), the background distortion image sequence obtained in step 3) and the flow field-free background speckle plate image sequence obtained in step 2) are processed to obtain a two-dimensional pixel offset, and then the modulus value of the speckle displacement field is calculated based on the two-dimensional pixel offset and the image is normalized and visualized to obtain a visualized image sequence. The background noise of the visualized image sequence is then reduced by combining the non-local means filtering algorithm (NLM) to obtain a high-precision and high-resolution pseudo-schlieren image sequence as follows:

[0017] Based on the cross-correlation algorithm, the two-dimensional pixel offset of the background distortion image relative to the flow-free background speckle plate image at each center pixel point (x, y) is calculated and represented by u(x, y) and v(x, y), respectively. Among them, u(x, y) represents the pixel offset of the center pixel point (x, y) in the horizontal direction, and v(x, y) represents the pixel offset of the center pixel point (x, y) in the vertical direction.

[0018] To construct a single-channel grayscale image, the modulus of the speckle displacement field is calculated based on the above two-dimensional pixel offset, as shown in the following formula:

[0019]

[0020] Wherein, ε is a micro-regularization term. In order to enhance the image contrast and meet the requirements of image processing, the image is normalized using the above modulus value. The calculation formula of the normalization is as follows:

[0021]

[0022] Among them, μ M and σ M are the mean and standard deviation of the modulus value M(x, y) in the entire image domain; then, the normalized image is linearly mapped to the standard 0-255 grayscale range to generate a visual image sequence I that can be directly displayed or used for subsequent processing. pseudo (i, j);

[0023] In order to further improve the quality of the pseudo-texture image and reduce the random noise caused by background inhomogeneity and imaging system, the non-local mean filtering algorithm is used to filter the above visualization image sequence I. pseudo (i, j) performs background noise reduction to obtain a high-precision and high-resolution pseudo-schlieren image sequence; the non-local mean filtering formula is as follows:

[0024]

[0025] Where Ω is the search neighborhood of the center pixel to be processed, C(x, y) is the normalization factor, w is the weight function, (x, y) is the coordinate of the center pixel; (i, j) is the coordinate of any pixel in the search neighborhood Ω.

[0026] In step 5), the pseudo-schlieren image sequence obtained in step 4) is combined with the real light path and the mapping relationship between the pixel offset on the camera imaging plane and the velocity vector in the flow field to calculate the flow field velocity field using the improved PIV-optical flow algorithm, thereby obtaining the final flow velocity of the flow field to be measured. The method is as follows:

[0027] According to the physical model of light propagation, light will be deflected due to the refractive index gradient when passing through the flow field. The horizontal deflection angle θ x and the vertical deflection angle θ y This results in a horizontal pixel offset δx' and a vertical pixel offset δy' on the camera imaging plane:

[0028] δx′=f·θ x +∈ x

[0029] δy′=f·θ y +∈ y

[0030] Where f is the focal length of high-speed camera 1, ∈ x ,∈ y To measure noise;

[0031] The horizontal pixel offset δx' and the vertical pixel offset δy' are related to the horizontal flow field velocity u flow and vertical flow velocity v flow The geometric relationship is expressed as:

[0032]

[0033] Where Za is the distance between the flow field to be measured and the high-speed camera, Δt is the image time interval, and K is the system calibration coefficient;

[0034] Finally, the improved PIV-optical flow algorithm is used to calculate the horizontal flow velocity u flow and vertical flow velocity v flow The flow field velocity field can be calculated by iterative solution, thereby obtaining the final flow velocity of the flow field to be measured. The formula is as follows:

[0035]

[0036] in, is the image spatial gradient, is the image time derivative, α is the smoothing term weight, β is the prior constraint weight, u p and v p is a priori estimate from the initial cross-correlation value.

[0037] The present invention has the following beneficial effects:

[0038] (1) It can realize non-contact full-field flow velocity measurement, overcoming the limitation of traditional PIV technology that relies on tracer particles; (2) The improved hybrid gradient operator and multi-scale feature fusion technology are used to greatly improve the calculation accuracy of the refractive index gradient field, reducing the uncertainty of flow velocity measurement to within 3.2%; (3) The system has a high degree of integration, and the measurement reliability is ensured through a standardized optical calibration process, and the complexity of optical components is reduced compared with traditional schlieren systems; (4) The pseudo-schlieren image combined with the PIV-optical flow fusion algorithm effectively solves the problem of insufficient spatial resolution in turbulent field measurement, and a reliable flow velocity distribution can be obtained under a 32×32 pixel window. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 Schematic diagram of the background schlieren experimental measurement system in the present invention.

[0040] Figure 2 is the speckle offset obtained after processing by the cross-correlation algorithm in the present invention.

[0041] Figure 3 It is the pseudo schlieren image in the present invention.

[0042] Figure 4 Flow chart of the flow field velocity measurement method using the background schlieren method based on pseudo schlieren images provided by the present invention. DETAILED DESCRIPTION

[0043] The background schlieren method for flow field velocity measurement based on pseudo schlieren images provided by the present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0044] like Figure 4 As shown, the background schlieren method flow field velocity measurement method based on pseudo schlieren image provided by the present invention includes the following steps performed in sequence:

[0045] 1) Build Figure 1 The background schlieren experimental measurement system shown in the figure comprises: a high-speed camera 1, a flow field to be measured 2, a background speckle plate 3, and an optical adjustment platform; wherein the lens of the high-speed camera 1 is directly facing the background speckle plate 3; the optical adjustment platform is arranged between the high-speed camera 1 and the background speckle plate 3; the flow field to be measured 2 is arranged on the optical adjustment platform and is located on the optical axis between the high-speed camera 1 and the background speckle plate 3;

[0046] In this embodiment, the high-speed camera 1 has a frame rate of 8000 frames per second and is equipped with a telecentric lens with a focal length of 100 mm. The distance between the high-speed camera 1 and the flow field to be measured 2 is 1000 mm, and the distance between the flow field to be measured 2 and the background speckle plate 3 is 800 mm. The adjustment accuracy of the optical adjustment platform is ±0.1 mm.

[0047] 2) Optically calibrate the background schlieren experimental measurement system. First, use a laser collimator and an optical adjustment platform to adjust the optical axis of the high-speed camera 1 to be perpendicular to the background speckle plate 3, with the angular deviation controlled within ±0.3°. Then, use a checkerboard calibration plate to calibrate the internal parameters of the high-speed camera 1. Then, under conditions without flow field disturbance, use the high-speed camera 1 to capture a reference image sequence of the background speckle plate 3. These reference image sequences are processed using PIVlab software to ensure that there is no obvious offset, thereby obtaining a flow-free background speckle plate image sequence.

[0048] In this embodiment, the background speckle plate 3 has a dot density of 600 dots / mm 2 Random dot pattern;

[0049] The method of using PIVlab software to process the reference image sequence is as follows: first, Gaussian filtering and histogram equalization are performed on the reference image sequence to enhance the contrast, and then lens distortion correction is performed based on the calibrated internal parameters to complete the preprocessing of the reference image sequence; then, the following method is used: Figure 2The cross-correlation algorithm of the multi-level iterative window shown in the figure uses fast Fourier transform (FFT) to accelerate the calculation, sets a 50% overlap rate, and uses the Gaussian three-point fitting method to accurately detect the correlation peak. Then, a normalized cross-correlation threshold (NCC>0.8) is used to filter out invalid vectors, and a local median filter is performed to eliminate outliers. Finally, the standard deviation of the full-field displacement vector is calculated (σ<0.05 pixels) to ensure a pass rate of more than 99.5%.

[0050] 3) First, start the flow field 2 to be measured. After the flow field 2 to be measured is stable, the high-speed camera 1 is synchronously triggered. The high-speed camera 1 starts to collect images of the background speckle plate 3 at a frame rate of 8000 frames per second. During the acquisition process, the exposure mode is used to record the image with a point density of 600 points / mm under the action of the flow field. 2 The distorted image sequence of the random dot pattern background speckle plate 3 is acquired, the signal-to-noise ratio and contrast of the distorted image are monitored in real time, and bad frame detection is performed immediately after the acquisition is completed, and finally a high temporal and spatial resolution background distorted image sequence that meets the requirements of subsequent processing is obtained and saved;

[0051] The exposure time of the high-speed camera 1 is 100 μs, the background distortion image is saved in a RAW lossless format, and the resolution of each frame of the background distortion image is 1024×680 pixels.

[0052] 4) Processing the background distortion image sequence obtained in step 3) and the flow field-free background speckle plate image sequence obtained in step 2) to obtain a two-dimensional pixel offset, then calculating the modulus of the speckle displacement field based on the two-dimensional pixel offset and performing normalized visualization processing on the images to obtain a visualized image sequence. Background noise reduction is then performed on the visualized image sequence using a non-local means filtering algorithm (NLM) to obtain a high-precision and high-resolution pseudo-schlieren image sequence;

[0053] Based on the cross-correlation algorithm, the two-dimensional pixel offset of the background distortion image relative to the flow-free background speckle plate image at each center pixel point (x, y) is calculated and represented by u(x, y) and v(x, y), respectively. Among them, u(x, y) represents the pixel offset of the center pixel point (x, y) in the horizontal direction, and v(x, y) represents the pixel offset of the center pixel point (x, y) in the vertical direction.

[0054] To construct a single-channel grayscale image, the modulus of the speckle displacement field is calculated based on the above two-dimensional pixel offset, as shown in the following formula:

[0055]

[0056] Where ε is a micro-regularization term that prevents instability in the calculation of division by zero or weak schlieren regions; this modulus value can be regarded as the intensity response of the local refractive index perturbation. In order to enhance the image contrast and meet the requirements of image processing, the image is normalized using the above modulus value. The calculation formula for the normalization is as follows:

[0057]

[0058] Among them, μ M and σ M are the mean and standard deviation of the modulus value M(x, y) in the entire image domain; then, the normalized image is linearly mapped to the standard 0-255 grayscale range to generate a visual image sequence I that can be directly displayed or used for subsequent processing. pseudo (i, j), this image can be regarded as a preliminary pseudo-schlieren image, and its grayscale change reflects the size of the speckle displacement;

[0059] In order to further improve the quality of the pseudo-texture image and reduce the random noise caused by background inhomogeneity and imaging system, the non-local mean filtering algorithm is used to filter the above visualization image sequence I. pseudo (i, j) performs background noise reduction and obtains Figure 3 The high-precision and high-resolution pseudo-schlieren image sequence shown in FIG. 1 ; the non-local mean filtering formula is as follows:

[0060]

[0061] Where Ω is the search neighborhood of the center pixel to be processed, C(x, y) is the normalization factor, w is the weight function, (x, y) is the coordinate of the center pixel; (i, j) is the coordinate of any pixel in the search neighborhood Ω.

[0062] 5) Based on the pseudo-schlieren image sequence obtained in step 4), combined with the real light path and the mapping relationship between the pixel offset on the camera imaging plane and the velocity vector in the flow field, the flow field velocity field is calculated using an improved PIV-optical flow algorithm to obtain the final flow velocity of the flow field to be measured;

[0063] According to the physical model of light propagation, light will be deflected due to the refractive index gradient when passing through the flow field. The horizontal deflection angle θ x and the vertical deflection angle θ y This results in a horizontal pixel offset δx' and a vertical pixel offset δy' on the camera imaging plane:

[0064] δx′=f·θ x +∈ x

[0065] δy′=f·θ y +∈ y

[0066] Where f is the focal length of high-speed camera 1, ∈ x ,∈ y To measure noise;

[0067] The horizontal pixel offset δx' and the vertical pixel offset δy' are related to the horizontal flow field velocity u flow and vertical flow velocity v flow The geometric relationship is expressed as:

[0068]

[0069] Wherein, Za is the distance between the flow field 2 to be measured and the high-speed camera 1, Δt is the image time interval, and K is the system calibration coefficient;

[0070] Finally, the improved PIV-optical flow algorithm is used to calculate the horizontal flow velocity u flow and vertical flow velocity v flow The flow field velocity field can be calculated by iterative solution, thereby obtaining the final flow velocity of the flow field 2 to be measured. The formula is as follows:

[0071]

[0072] in, is the image spatial gradient, is the image time derivative, α is the smoothing term weight, β is the prior constraint weight, u p and v p is a priori estimate from the initial cross-correlation value.

Claims

1. A method for measuring flow velocity in a flow field using a background schlieren method based on pseudo-schlieren images, characterized in that: The method for measuring flow field velocity using a background schlieren method using a pseudo schlieren image comprises the following steps performed in sequence: 1) A background schlieren experimental measurement system is constructed, the system comprising: a high-speed camera (1), a flow field to be measured (2), a background speckle plate (3) and an optical adjustment platform; wherein the lens of the high-speed camera (1) faces the background speckle plate (3); the optical adjustment platform is arranged between the high-speed camera (1) and the background speckle plate (3); the flow field to be measured (2) is arranged on the optical adjustment platform and is located on the optical axis of the high-speed camera (1) and the background speckle plate (3); 2) The above-mentioned background schlieren experimental measurement system is optically calibrated. First, the optical axis of the high-speed camera (1) is adjusted to be perpendicular to the background speckle plate (3) by using a laser collimator and an optical adjustment platform. Then, the internal parameters of the high-speed camera (1) are calibrated using a checkerboard calibration plate. After that, under the condition of no flow field disturbance, the high-speed camera (1) is used to collect a reference image sequence of the background speckle plate (3). The reference image sequence is processed using PIVlab software to ensure that there is no obvious offset, thereby obtaining a flow-free background speckle plate image sequence. 3) First, the flow field to be measured (2) is started, and after the flow field to be measured (2) is stabilized, the high-speed camera (1) is synchronously triggered, and the image of the background speckle plate (3) is started to be collected by the high-speed camera (1). During the collection process, the distorted image sequence of the background speckle plate (3) under the action of the flow field is recorded in an exposure mode, and the signal-to-noise ratio and contrast of the distorted image are monitored in real time. After the collection is completed, bad frame detection is immediately performed, and finally a background distorted image sequence with high temporal and spatial resolution that meets the requirements of subsequent processing is obtained and saved; 4) processing the background distortion image sequence obtained in step 3) and the flow field-free background speckle plate image sequence obtained in step 2) to obtain a two-dimensional pixel offset, then calculating the modulus of the speckle displacement field based on the two-dimensional pixel offset and performing normalized visualization processing on the images to obtain a visualized image sequence, then performing background noise reduction on the visualized image sequence using a non-local means filtering algorithm to obtain a high-precision and high-resolution pseudo-schlieren image sequence; 5) Based on the pseudo-schlieren image sequence obtained in step 4), combined with the real light path and the mapping relationship between the pixel offset on the camera imaging plane and the velocity vector in the flow field, the flow field velocity field is calculated using the improved PIV-optical flow algorithm to obtain the final flow velocity of the flow field to be measured.

2. The method for measuring flow field velocity using the background schlieren method based on pseudo schlieren images according to claim 1, characterized in that: In step 1), the high-speed camera (1) has a frame rate of 8000 frames per second and is equipped with a telecentric lens with a focal length of 100 mm; the distance between the high-speed camera (1) and the flow field to be measured (2) is 1000 mm, and the distance between the flow field to be measured (2) and the background speckle plate (3) is 800 mm; and the adjustment accuracy of the optical adjustment platform is ±0.1 mm.

3. The method for measuring flow field velocity using the background schlieren method based on pseudo schlieren images according to claim 1, characterized in that: In step 2), the background speckle plate (3) has a dot density of 600 dots / mm 2 Random dot pattern; The method for processing the reference image sequence using PIVlab software is as follows: first, the reference image sequence is subjected to Gaussian filtering for noise reduction and histogram equalization to enhance contrast, and then lens distortion correction is performed based on calibrated internal parameters, thereby completing the preprocessing of the reference image sequence; then, a multi-level iterative window cross-correlation algorithm is used, the calculation is accelerated by fast Fourier transform, a 50% overlap rate is set, and a Gaussian three-point fitting method is used to accurately detect correlation peaks; Then, the normalized cross-correlation threshold NCC is used to filter out invalid vectors, and local median filtering is performed to eliminate outliers. Finally, the standard deviation σ of the full-field displacement vector is calculated to ensure that the pass rate reaches more than 99.5%; the normalized cross-correlation threshold NCC>0.8, and the standard deviation σ<0.05 pixels.

4. The method for measuring flow field velocity using the background schlieren method based on pseudo schlieren images according to claim 1, characterized in that: In step 3), the exposure time of the high-speed camera (1) is 100 μs, the background distortion image is saved in a RAW lossless format, and the resolution of each frame of the background distortion image is 1024×680 pixels.

5. The method for measuring flow field velocity using the background schlieren method based on pseudo schlieren images according to claim 1, characterized in that: In step 4), the background distortion image sequence obtained in step 3) and the flow field-free background speckle plate image sequence obtained in step 2) are processed to obtain a two-dimensional pixel offset, and then the modulus value of the speckle displacement field is calculated based on the two-dimensional pixel offset and the image is normalized and visualized to obtain a visualized image sequence. The background noise of the visualized image sequence is then reduced by combining the non-local means filtering algorithm to obtain a high-precision and high-resolution pseudo-schlieren image sequence as follows: Based on the cross-correlation algorithm, the two-dimensional pixel offset of the background distortion image relative to the flow-free background speckle plate image at each center pixel point (x, y) is calculated and represented by u(x, y) and v(x, y), respectively. Among them, u(x, y) represents the pixel offset of the center pixel point (x, y) in the horizontal direction, and v(x, y) represents the pixel offset of the center pixel point (x, y) in the vertical direction. To construct a single-channel grayscale image, the modulus of the speckle displacement field is calculated based on the above two-dimensional pixel offset, as shown in the following formula: Wherein, ε is a micro-regularization term. In order to enhance the image contrast and meet the requirements of image processing, the image is normalized using the above modulus value. The calculation formula of the normalization is as follows: Among them, μ M and σ M are the mean and standard deviation of the modulus value M(x, y) in the entire image domain; then, the normalized image is linearly mapped to the standard 0-255 grayscale range to generate a visual image sequence I that can be directly displayed or used for subsequent processing. pseudo (i, j); In order to further improve the quality of the pseudo-texture image and reduce the random noise caused by background inhomogeneity and imaging system, the non-local mean filtering algorithm is used to filter the above visualization image sequence I. pseudo (i, j) performs background noise reduction to obtain a high-precision and high-resolution pseudo-schlieren image sequence; the non-local mean filtering formula is as follows: Where Ω is the search neighborhood of the center pixel to be processed, C(x, y) is the normalization factor, w is the weight function, (x, y) is the coordinate of the center pixel; (i, j) is the coordinate of any pixel in the search neighborhood Ω.

6. The method for measuring flow field velocity using the background schlieren method based on pseudo schlieren images according to claim 1, characterized in that: In step 5), the pseudo-schlieren image sequence obtained in step 4) is combined with the real light path and the mapping relationship between the pixel offset on the camera imaging plane and the velocity vector in the flow field to calculate the flow field velocity field using the improved PIV-optical flow algorithm, thereby obtaining the final flow velocity of the flow field to be measured. The method is as follows: According to the physical model of light propagation, light will be deflected due to the refractive index gradient when passing through the flow field, and the horizontal deflection angle θ x and the vertical deflection angle θ y This results in a horizontal pixel offset δx' and a vertical pixel offset δy' on the camera imaging plane: δx′=f·θ x +∈ x δy′=f·θ y +∈ y Where, f is the focal length of the high-speed camera (1), ∈ x ,∈ y To measure noise; The horizontal pixel offset δx' and the vertical pixel offset δy' are related to the horizontal flow field velocity u flow and vertical flow velocity v flow The geometric relationship is expressed as: Wherein, Za is the distance between the flow field to be measured (2) and the high-speed camera (1), Δt is the image time interval, and K is the system calibration coefficient; Finally, the improved PIV-optical flow algorithm is used to calculate the horizontal flow velocity u flow and vertical flow velocity v flow The flow field velocity field can be calculated by iterative solution, thereby obtaining the final flow velocity of the flow field to be measured (2), and the formula is as follows: in, is the image spatial gradient, is the image time derivative, α is the smoothing term weight, β is the prior constraint weight, u p and v p is a priori estimate from the initial cross-correlation value.

Images