Shock wave diffusion process testing method based on double cameras and digital schlieren

Abstract

The invention discloses a shock wave diffusion process testing method based on double cameras and digital schlieren. The method comprises the steps that two high-speed cameras are used for shooting digital schlieren images generated in the shock wave diffusion process; performing image splicing on the acquired images shot by the double cameras; the position of the shock wave is calculated through the spliced image, shock wave position data at different moments are obtained, and the speed of shock wave propagation in the x-axis direction and the y-axis direction is obtained. Based on binocular vision, a digital schlieren system is adopted to collect an air flow field information image containing a shock wave propagation process; splicing of double-camera images is achieved through an improved homography matrix method, the shock wave testing range is expanded, the whole shock wave propagation process is successfully captured, the shock wave diffusion testing accuracy is improved, and visual monitoring of wave types represented by shock waves in various transparent media can be achieved.

Description

Technical Field

[0001] This invention relates to the field of shock wave testing technology, and specifically to a shock wave diffusion process testing method based on dual cameras and digital schlieren imaging. Background Technology

[0002] Shock wave overpressure testing is a common requirement in various safety tests and assessments, including those involving sympathetic explosions. The conventional measurement method involves placing overpressure sensors around the test specimen, centered on the explosion epicenter, from near to far. After the explosion, the shock wave propagates through the pressure-sensing surfaces of the sensors, recording the overpressure value at each sensor's location at that moment. The data is then collected after the test, and the shock wave overpressure data field is obtained by fitting the collected overpressure values ​​from various points. Currently, schlieren technology is used as an auxiliary measurement method for testing the diffusion process and velocity of shock waves. This involves observing the propagation of the shock wave front in the acquired video and image data. Therefore, analyzing and extracting data from high-speed photography systems for shock wave overpressure testing provides excellent data conditions.

[0003] Schlieren technique, also known as the Schlieren method, is a classic optical display technique. Its basic principle is based on the fact that the refractive index gradient of light in the measured flow field is proportional to the air density of the flow field. It transforms the change in density gradient in the flow field into a change in relative light intensity on the recording plane, making areas with drastic density changes, such as shock waves and compression waves in compressible flow fields, observable and resolvable images, and thus recording them.

[0004] Schlieren velocimetry works by using a high-speed camera to capture schlieren images of a flow. When the camera's frame rate is high enough, the movement of turbulent structures or flow interfaces in the flow field can be captured. Based on the changes in refractive index caused by the flow in the test flow field, the velocity of the medium in the test flow field is calculated. Schlieren velocimetry can accurately capture and present subtle changes in refractive index in space, clearly showing transparent phenomena that are originally difficult to present intuitively, such as the distribution of atmospheric temperature, heat, and pressure, as well as changes in gas flow. At the same time, it can detect foreign objects in transparent media, and has important application value in scientific research, industrial production, and many other fields.

[0005] The existing technology, disclosed in CN114858972A, describes a method and apparatus for measuring the back parameters of an explosion shock wave based on background schlieren technology. Its technical solution includes acquiring parameters such as shock wave back density, pressure, velocity, temperature, and Mach number based on background schlieren technology. However, it is not suitable for observing long-distance, large-scale shock waves, as is the case in the missile field. Using multiple cameras can overcome the difficulty of observing long-distance shock waves with a single camera, but it also brings the challenge of how to conduct tests based on images acquired by two cameras. Summary of the Invention

[0006] 1. The technical problem to be solved:

[0007] To address the aforementioned technical problems, this invention provides a method for testing the shock wave diffusion process based on dual cameras and digital schlieren imaging.

[0008] 2. Technical Solution:

[0009] A method for testing the shock wave diffusion process based on dual cameras and digital schlieren imaging, characterized by comprising:

[0010] Step 1: Use two high-speed cameras to capture digital schlieren images of the shock wave propagation process;

[0011] Step 2: Stitch the images captured by the dual cameras;

[0012] Step 3: Calculate the location S of the shock wave using the stitched image. * :(x i ,y i );

[0013] Step 4: Obtain shock wave position data at different times to obtain the shock wave propagation velocity V in the x and y axis directions: (x i ,y i ).

[0014] Furthermore, in step one, a digital schlieren system is used to acquire digital schlieren images; the digital schlieren system includes two high-speed cameras, a background panel, and a hot airflow generator.

[0015] The background plate is laid along the direction of shock wave propagation, and its length is sufficient to capture the entire propagation process of the shock wave; the upper surface of the background plate is provided with speckle patterns for locating the shock wave, and the speckle size is calculated using the following formula:

[0016]

[0017] In the above formula, α represents the resolution of the high-speed camera; e(v f ) is affected by wind speed v f Error coefficient influenced by external factors; A represents the speckle area size, (x i ,y i L*:(V) represents the coordinate position of the i-th speckle; i ,t) represents the distance from the i-th spot to the camera, which is related to the shock wave propagation distance; t i The time it takes for the shock wave to propagate to the i-th speckle;

[0018] The background plate has an encoding point set in the middle for later image stitching and reconstruction. The size of the encoding point is as follows:

[0019]

[0020] In the above formula, B represents the area of ​​the coding point; The smaller the value of this parameter, the more obvious the strange image becomes.

[0021] The hot air flow generating device uses laser-induced heating to generate hot air flow, and achieves the creation and generation of complex hot air flow patterns by dynamically controlling the heating position;

[0022] The two high-speed cameras are respectively positioned on the left and right sides of the background panel, with the cameras perpendicular to the background panel.

[0023] Furthermore, step two specifically includes:

[0024] S21: Acquire digital schlieren images from two cameras at the same time, extract feature points from each image, obtain the perspective matrix after registration, and use perspective transformation to convert the two images into images with the same degree of curvature, as follows:

[0025]

[0026] In the above formula, and Let be the perspective matrix, which represents the linear mapping of the two images respectively; This indicates that the coordinates in the stitched image are (x, y, y). i， y i Data information; k = sinα, α is the angle between the vertical lines of each image. The smaller the value of α, the more obvious the singular peaks. However, as α decreases, the image information also becomes more obvious. Therefore, the value of α should be selected according to the actual situation. These represent the data information of the same coordinate points in the two images respectively;

[0027] S22: The images obtained in step S21 are quickly stitched together using an improved homography matrix; specifically:

[0028] S221: Divide the converted image into multiple planar regions, estimate the local homography of each planar region, and combine Delauany triangulation to smooth the transition boundary;

[0029] S222: For each planar region, use K-means clustering to identify the plane to which the feature points belong, and fit multiple homography matrices corresponding to multiple planar regions;

[0030] S223: Correct image errors using depth maps or sparse 3D point clouds, and optimize the homography matrix using depth-weighted image error calculation.

[0031] S224: For the homography matrix of each planar region, use Poisson mixing combined with masking to transition regions with different weighted homography matrices to obtain an improved homography matrix for the entire image; based on the improved homography matrix, map all region images to the same coordinate system and stitch them into an image.

[0032] Furthermore, step three specifically includes:

[0033] S31: Obtain information from the stitched image and perform density field reconstruction:

[0034]

[0035] in It is the dielectric constant; ρ is the gray-level gradient of the schlieren image; n0 is the refractive index of the environment; ρ(x,y) represents the air density field. This represents the square of the density field gradient;

[0036] S32: Calculate the density field gradient magnitude of the stitched image using the following formula:

[0037]

[0038] S33: Determine the location of the shock wave. * :(x i ,y i ):

[0039]

[0040] Where H k Represents the discriminant coefficients related to the density field;

[0041] Get the number of frames Z captured by the camera per unit time. i Calculate the velocity V of the shock wave propagating in the x and y directions: (x i ,y i ) = S * :(x i ,y i )÷Z i (7)

[0042] Z i This represents the number of frames in the image.

[0043] 3. Beneficial effects:

[0044] This method provides a shock wave diffusion process testing method based on dual cameras and digital schlieren imaging. Based on binocular vision, a digital schlieren system is used to acquire airflow field information images containing the shock wave propagation process. By using an improved homography matrix method to stitch the dual camera images, the testing range of shock waves is expanded, and the entire shock wave propagation process is successfully captured, improving the accuracy of shock wave diffusion testing. This method can realize the visual monitoring of wave types represented by shock waves in various transparent media. Attached Figure Description

[0045] Figure 1 This is the overall flowchart of this method;

[0046] Figure 2 This is a schematic diagram of the digital schlieren system of this method;

[0047] Figure 3 To verify the density field display at a certain moment in the example;

[0048] Figure 4 The front view of the background board for verification examples;

[0049] Figure 5 To verify the time-varying curve of the shock wave position in the X direction obtained in the example.

[0050] Figure labels: 1. High-speed camera; 2. Background board; 3. Hot airflow generator; 4. High-frequency white LED. Detailed Implementation

[0051] The present invention will now be described in detail with reference to the accompanying drawings.

[0052] As attached Figure 1 As shown, a shock wave diffusion process testing method based on dual cameras and digital schlieren imaging is characterized by comprising:

[0053] Step 1: Use two high-speed cameras to capture digital schlieren images of the shock wave propagation process;

[0054] Step 2: Stitch the images captured by the dual cameras;

[0055] Step 3: Calculate the location S of the shock wave using the stitched image. * :(x i ,y i );

[0056] Step 4: Obtain shock wave position data at different times to obtain the shock wave propagation velocity V in the x and y axis directions: (x i ,y i ).

[0057] Furthermore, in step one, a digital schlieren system is used to acquire digital schlieren images; the digital schlieren system includes two high-speed cameras 1, a background plate 2, and a hot airflow generator 3;

[0058] The background plate is laid along the direction of shock wave propagation, and its length is sufficient to capture the entire propagation process of the shock wave; the upper surface of the background plate is provided with speckle patterns for locating the shock wave, and the speckle size is calculated using the following formula:

[0059]

[0060] In the above formula, α represents the resolution of the high-speed camera; e(v f ) is affected by wind speed v f Error coefficient influenced by external factors; A represents the speckle area size, (x i ,y i L*:(V) represents the coordinate position of the i-th speckle; i ,t) represents the distance from the i-th spot to the camera, which is related to the shock wave propagation distance; t i The time it takes for the shock wave to propagate to the i-th speckle;

[0061] The background plate has an encoding point set in the middle for later image stitching and reconstruction. The size of the encoding point is as follows:

[0062] In the above formula, B represents the area of ​​the coding point; The smaller the value of the preset proportional parameter, the more obvious the strange image; the hot air flow generating device uses laser-induced heating to generate hot air flow, and realizes the creation and generation of complex hot air flow patterns by dynamically controlling the heating position;

[0063] The two high-speed cameras are respectively positioned on the left and right sides of the background panel, with the cameras perpendicular to the background panel.

[0064] Furthermore, step two specifically includes:

[0065] S21: Acquire digital schlieren images from two cameras at the same time, extract feature points from each image, obtain the perspective matrix after registration, and use perspective transformation to convert the two images into images with the same degree of curvature, as follows:

[0066]

[0067] In the above formula, and Let be the perspective matrix, which represents the linear mapping of the two images respectively; This indicates that the coordinates in the stitched image are (x, y, y). i， y iData information; k = sinα, α is the angle between the vertical lines of each image. The smaller the value of α, the more obvious the singular peaks. However, as α decreases, the image information also becomes more obvious. Therefore, the value of α should be selected according to the actual situation. These represent the data information of the same coordinate points in the two images respectively;

[0068] S22: The images obtained in step S21 are quickly stitched together using an improved homography matrix; specifically:

[0069] S221: Divide the converted image into multiple planar regions, estimate the local homography of each planar region, and combine Delauany triangulation to smooth the transition boundary;

[0070] S222: For each planar region, use K-means clustering to identify the plane to which the feature points belong, and fit multiple homography matrices corresponding to multiple planar regions;

[0071] S223: Correct image errors using depth maps or sparse 3D point clouds, and optimize the homography matrix using depth-weighted image error calculation.

[0072] S224: For the homography matrix of each planar region, use Poisson mixing combined with masking to transition regions with different weighted homography matrices to obtain an improved homography matrix for the entire image; based on the improved homography matrix, map all region images to the same coordinate system and stitch them into an image.

[0073] Furthermore, step three specifically includes:

[0074] S31: Obtain information from the stitched image and perform density field reconstruction:

[0075]

[0076] in It is the dielectric constant; ρ is the gray-level gradient of the schlieren image; n0 is the refractive index of the environment; ρ(x,y) represents the air density field. This represents the square of the density field gradient;

[0077] S32: Calculate the density field gradient magnitude of the stitched image using the following formula:

[0078]

[0079] S33: Determine the location of the shock wave. * :(x i ,y i ):

[0080]

[0081] Where H k Represents the discriminant coefficients related to the density field;

[0082] Get the number of frames Z captured by the camera per unit time. i Calculate the velocity V of the shock wave propagating in the x and y directions: (x i ,y i ) = S * :(x i ,y i )÷Z i (7)

[0083] Z i This represents the number of frames in the image.

[0084] Verification example:

[0085] This verification example demonstrates the accuracy of the proposed solution by constructing a digital schlieren system.

[0086] First, set up a digital schlieren system, as shown in the attached image. Figure 2 As shown, the digital schlieren system includes two high-speed cameras 1, a background 2, a shock wave 3, and a high-frequency white LED 4; the camera parameters are 5 megapixel resolution and a sampling frequency of 30Hz; the background light is selected from the high-frequency white LED 4; the background 2 is 10m*3m in size; the camera is perpendicular to the background wall, and the height of both the camera and the light source is 1.5m. The centers of the two cameras are aligned with the background wall at positions 2.5m and 7.5m from left to right, respectively, with a distance of 5m between the two cameras and a distance of 3-5m from the light source 4 to the background 2. If a 100mm lens is used, the distance between the camera and the background 2 is approximately 31.25m, and the distance between the shock wave and the background 3 is approximately 10m. The overlapping part of the two camera images (located at a position of 5m from left to right on the background 2) is marked with coded points for image stitching. The size of the coded points is determined based on the above data using formula (2); at the same time, the background 2 is made into a speckle background using a speckle preparation tool, and its size is obtained based on formula (1). A portable laptop is connected to the two high-speed cameras to acquire images, and the system is triggered by a synchronization trigger. See attached Figure 3 The image shown is a density field display at a certain moment obtained in the verification example; as attached. Figure 4 This is a front view of the background panel for this verification example; Figure 5 This is a graph showing the time-varying position of the shock wave in the X direction obtained in this verification example.

[0087] Accuracy analysis:

[0088] In this verification example, the overall error of target feature point extraction is 0.16 pixels (including 0.15 pixels due to camera lens distortion, 0.02 pixels due to digital image correlation algorithm matching error, 0.15 pixels due to image stitching error between two cameras, and 0.03 pixels due to image point extraction error caused by high-speed motion blur).

[0089] Position test accuracy: The displacement error of the shock wave in the X and Y directions is approximately 3.1mm * 0.16 = 0.496mm. Since the speed of the shock wave is greater than 340m / s, the displacement of the shock wave between adjacent image frames is greater than 340m / s / 3200fps = 106mm. Therefore, the position resolution error of the shock wave is less than 0.496mm / 106mm = 0.46%.

[0090] Velocity test accuracy: The velocity measurement resolution of the shock wave between adjacent image frames is 0.496mm*3200fps=1.59m / s, and the velocity resolution error of the shock wave is less than 1.59 / 340=0.05%.

[0091] Although the present invention has been disclosed above with reference to preferred embodiments, these are not intended to limit the invention. Any person skilled in the art can make various changes or modifications without departing from the spirit and scope of the invention. Therefore, the scope of protection of the present invention should be defined by the scope of the claims of this application.

Claims

1. A method for testing the shock wave propagation process based on a dual camera and digital schlieren, characterized in that: The application relates to a method for obtaining a shock wave propagation image. The method comprises the following steps: Step one: acquiring digital schlieren images generated in the process of shock wave propagation by using two high-speed cameras; Step three: Calculate the position of the shock wave S from the stitched image * :(x i ,y i ) Step four: Obtain the shock wave position data at different time to get the speed V of the shock wave propagation in the x, y axis direction: i (x i ,y ) 2. The method according to claim 1, wherein the method is characterized by: Step two: image splicing of the images acquired by the two cameras; In step one, a digital schlieren system is used to acquire the digital schlieren images; the digital schlieren system comprises two high-speed cameras, a background plate and a hot air flow generating device. In the above formula, a is the resolution of the high-speed camera; e(v f ) is the error coefficient affected by the wind speed v f ; A represents the size of the speckle area, (x i , y i ) represents the coordinate position of the i-th speckle; L*(v i , t) represents the distance from the i-th speckle to the camera, which is the propagation distance of the shock wave; t i is the time for the shock wave to propagate to the i-th speckle; The background plate is laid along the direction of shock wave propagation, and the length of the background plate can realize the capture of the whole process of shock wave propagation; the upper surface of the background plate is provided with speckles for positioning the shock wave, and the size of the speckles is obtained by the following formula: In the above formula, B represents the area size of the encoding point; The preset proportion parameter is smaller, and the singular image is more obvious; the hot air flow generating device adopts a laser-induced heating mode to generate hot air flow, and through dynamic control of the heating position, a complex hot flow pattern is manufactured and generated. An encoding point is arranged at the middle position of the background plate for image splicing and reconstruction in the later stage, and the size of the encoding point is as follows:

3. The method according to claim 2, wherein: The two high-speed cameras are arranged on the left and right sides of the background plate respectively, and the cameras are perpendicular to the background plate. Step two specifically comprises the following steps: In the above formula, and are perspective matrices respectively linear mapping of two images; represents the data information of the coordinate (x i， y i ) in the spliced image; k = sin α, α is the included angle between the perpendicular lines of each image, the smaller the α value, the more obvious the singular peak value, but as α decreases, the image information will also become obvious, so the α value should be selected according to the actual situation; respectively represent the data information of the same coordinate points of two images; S21: acquiring the digital schlieren images collected by the two cameras at the same time, extracting feature points from each image, obtaining a perspective matrix after registration, and converting the two images into images with the same bending degree by using perspective transformation, specifically as follows: S22: quickly splicing the images obtained in step S21 by using an improved homography matrix; specifically as follows: S221: dividing the converted image into multiple planar regions, estimating local homography for each planar region, and combining Delauany triangular partition to smooth the transition boundary; S222: using K-means clustering to cluster the feature points belonging to the planar region for each planar region, and fitting multiple homography matrices corresponding to the multiple planar regions; S223: correcting image errors by using a depth map or a sparse 3D point cloud, and optimizing the homography matrix by using image error depth weighting; 4. The method according to claim 3, wherein: S224: using Poisson mixing combined with a mask to transition different weighted homography matrix regions to obtain an improved homography matrix of the whole image; and mapping all the region images to the same coordinate system according to the improved homography matrix to splice them into one image. Step three specifically comprises the following steps: wherein is a medium constant; is a schlieren image gray scale gradient; n0is an ambient refractive index; p(x,y) represents an air density field; represents the square of the density field gradient; S31: acquiring the information of the spliced image, and performing density field reconstruction: S32: calculating the density field gradient amplitude of the spliced image according to the following formula: S33: Determine the Shockwave position S * :(x i ,y i ) where H k denotes the discriminant coefficient related to the density field; Get the number of frames Z captured by the camera per unit time. i Calculate the velocity V of the shock wave propagating in the x and y directions: (x i ,y i ) = S * :(x i ,y i )÷Z i (7) where Z i is the number of frames in the image.

Images