Flow field measurement region division method, device, storage medium, and program product

Abstract

This application provides a method, device, storage medium, and program product for segmenting a flow field measurement region, relating to the field of fluid measurement technology. The method includes: acquiring particle concentration information based on the spatial location of particles in a flow field; segmenting the flow field into regions based on the particle concentration information according to a boundary concentration threshold corresponding to the flow field, obtaining a first boundary between high-particle-concentration regions and low-particle-concentration regions in the flow field; performing morphological correction on the first boundary to obtain a second boundary; and determining the flow field measurement region segmentation result based on the second boundary. This application can improve the accuracy of flow field measurement region segmentation and enhance the accuracy of flow field measurement.

Description

Technical Field

[0001] This application relates to the field of fluid measurement technology, and in particular to a method, device, storage medium and program product for segmenting a flow field measurement region. Background Technology

[0002] In common industrial equipment such as various rotating machinery, heat exchangers, and fluid transport systems, the internal fluids exhibit certain multidimensional characteristics. For these industrial devices, the fluid state directly affects the overall performance and dynamic stability of the system. As industrial equipment develops towards higher operating parameters, wider operating ranges, and stronger dynamic response capabilities, the fluid state exhibits significant multidimensional and unsteady characteristics. Accurate flow field measurement of flow systems, achieving full-domain precise capture of complex flow structures and their evolution mechanisms, can provide crucial data support for a deeper understanding of the nature of flow, verification of numerical models, and optimization of system performance.

[0003] In related technologies, the measurement area is divided according to the spatial particle concentration distribution and concentration threshold in the fluid, and different measurement methods are used in different particle concentration areas to achieve zoned measurement of the flow field. However, this method suffers from inaccurate measurement area segmentation and poor flow field measurement results. Summary of the Invention

[0004] This application provides a method, device, storage medium, and program product for segmenting a flow field measurement region, which can improve the accuracy of flow field measurement region segmentation and improve the accuracy of flow field measurement.

[0005] In a first aspect, embodiments of this application provide a method for segmenting a flow field measurement region, including:

[0006] Obtain particle concentration information based on the spatial location of particles in the flow field;

[0007] Based on the boundary concentration threshold corresponding to the flow field, the flow field is divided into regions according to the particle concentration information to obtain the first boundary between the high particle concentration region and the low particle concentration region in the flow field.

[0008] The first boundary is morphologically corrected to obtain the second boundary;

[0009] Based on the second boundary, the flow field measurement region segmentation result is determined.

[0010] In one implementation, morphological correction is performed on the first boundary to obtain the second boundary, including:

[0011] Merging adjacent high-particle-concentration regions in the flow field yields a high-particle-concentration connected region, and merging adjacent low-particle-concentration regions in the flow field yields a low-particle-concentration connected region.

[0012] In high-particle-concentration or low-particle-concentration connected regions, connected regions with a volume smaller than a volume threshold are identified as invalid connected regions.

[0013] Delete the region corresponding to the invalid connected region from the high particle concentration region or the low particle concentration region to obtain the initial corrected boundary between the high particle concentration region and the low particle concentration region.

[0014] The second boundary is obtained based on the initial corrected boundary.

[0015] In one implementation, obtaining the second boundary based on the initial correction boundary includes:

[0016] Morphological opening and morphological closing operations are performed on the high particle concentration region and the low particle concentration region respectively to obtain the modified boundary that morphologically modifies the first boundary.

[0017] Median filtering is applied to the corrected boundary to obtain the second boundary.

[0018] In one implementation, determining the flow field measurement region segmentation result based on the second boundary includes:

[0019] Based on the average particle concentration in the high-particle-concentration region and the low-particle-concentration region divided by the second boundary, determine the particle concentration ratio between the high-particle-concentration region and the low-particle-concentration region.

[0020] Based on the correlation between the particle concentration ratio and the boundary concentration threshold, the boundary concentration threshold associated with the particle concentration ratio is updated to the boundary concentration threshold corresponding to the flow field.

[0021] Based on the boundary concentration threshold corresponding to the flow field, the second boundary is iteratively optimized to obtain the flow field measurement region segmentation result. The iterative optimization process includes correcting and optimizing the second boundary layer by layer outward from the region divided by the second boundary, and determining whether the average particle concentration of the corrected region satisfies the relationship between the corresponding region and the boundary concentration threshold. If it satisfies, the flow field measurement region segmentation result is obtained; if it does not satisfy, the second boundary is corrected and optimized layer by layer outward from the region divided by the second boundary.

[0022] In one implementation, obtaining particle concentration information at the spatial location of particles in the flow field includes:

[0023] Image calibration establishes a mapping relationship between the spatial location of the particle and the pixel space of the grayscale image, thereby obtaining the grayscale image corresponding to the spatial location of the particle.

[0024] Based on multiplicative algebraic reconstruction technology, a gray volume corresponding to a spatial location is constructed from a grayscale image;

[0025] In a grayscale volume, adjacent voxels with similar grayscale values ​​are merged to obtain the particles corresponding to the voxels.

[0026] Based on the particles, obtain the particle concentration information of the spatial location of the particles in the flow field.

[0027] In one implementation, obtaining particle concentration information at the spatial location of particles in the flow field includes:

[0028] Establish the Vino polyhedron corresponding to the particle;

[0029] The local particle concentration is determined by the reciprocal of the volume of the Vino polyhedron;

[0030] The local particle concentration is interpolated to spatial grid points corresponding to the particle physical coordinates using linear interpolation.

[0031] The local particle concentration at the spatial grid points is smoothed to obtain the particle concentration information at the spatial location of the particles.

[0032] In one implementation, establishing the Vino polyhedron corresponding to the particle includes:

[0033] The particle is translated along a preset direction to obtain the translated particle.

[0034] Centered on the particle, construct an extended spatial point set based on the particle and its translation;

[0035] Generate the Vino polyhedron for the extended space point set and establish the Vino polyhedron corresponding to the particle.

[0036] Secondly, embodiments of this application provide a flow field measurement region segmentation device, comprising:

[0037] The acquisition module is used to acquire particle concentration information at the spatial location of particles in the flow field;

[0038] The segmentation module is used to segment the flow field based on the boundary concentration threshold corresponding to the flow field and according to the particle concentration information to obtain the first boundary between the high particle concentration region and the low particle concentration region in the flow field.

[0039] The correction module is used to perform morphological correction on the first boundary to obtain the second boundary;

[0040] The determination module is used to determine the flow field measurement region segmentation result based on the second boundary.

[0041] Thirdly, embodiments of this application provide a flow field measurement region segmentation device, including: a memory and a processor;

[0042] The memory stores the instructions that the computer executes;

[0043] The processor executes computer execution instructions stored in memory, causing the processor to perform the methods described in the various possible implementations of the first aspect above.

[0044] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed, are used to implement the methods described in the various possible implementations of the first aspect above.

[0045] Fifthly, embodiments of this application provide a computer program product, including a computer program that, when executed, implements the methods described in the various possible implementations of the first aspect above.

[0046] The flow field measurement region segmentation method, device, storage medium, and program product provided in this application obtain particle concentration information of the spatial location of particles in the flow field; based on the boundary concentration threshold corresponding to the flow field, the flow field is segmented according to the particle concentration information to obtain a first boundary between high particle concentration regions and low particle concentration regions in the flow field; morphological correction is performed on the first boundary to obtain a second boundary; and the flow field measurement region segmentation result is determined based on the second boundary. By performing morphological correction on the first boundary to eliminate burrs and pits in the first boundary, the obtained second boundary is smoother and more accurate, making the flow field measurement region segmentation result determined based on the second boundary more accurate, providing an accurate basis for subsequent zonal measurement of the flow field, and improving the accuracy of flow field measurement. Attached Figure Description

[0047] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0048] Figure 1 A flowchart illustrating the flow field measurement region segmentation method provided in this application embodiment;

[0049] Figure 2 A schematic diagram of the particle concentration acquisition process provided in an embodiment of this application;

[0050] Figure 3 This is a schematic diagram illustrating the change in concentration at the region boundary as a function of particle concentration ratio, provided in an embodiment of this application.

[0051] Figure 4 The curve showing the change of normalized concentration value at the interface with particle concentration ratio provided in the embodiments of this application;

[0052] Figure 5 This is a schematic diagram of the dynamic threshold method for identifying particle boundaries provided in an embodiment of this application;

[0053] Figure 6 This is a schematic diagram of the Vino polyhedron generation process provided in an embodiment of this application;

[0054] Figure 7 A schematic diagram of the Vino polyhedron provided for an embodiment of this application. Figure 1 ;

[0055] Figure 8 This is a schematic diagram of a flow field measurement region segmentation scenario provided in an embodiment of this application;

[0056] Figure 9 A schematic diagram of the Vino polyhedron provided for an embodiment of this application. Figure 2

[0057] Figure 10 This is a schematic diagram of the recognition region after median filtering provided in an embodiment of this application;

[0058] Figure 11 A schematic diagram showing the identification region and its comparison with the theoretical region provided in the embodiments of this application;

[0059] Figure 12 A schematic diagram of the flow field measurement region segmentation device provided in the embodiments of this application;

[0060] Figure 13 A schematic diagram of the flow field measurement region segmentation device provided in the embodiments of this application.

[0061] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0062] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0063] Currently, experimental research methods for multidimensional non-flow can be mainly divided into two categories: invasive measurement and non-invasive measurement. Invasive measurement methods include pressure probes and hot-wire / hot-film anemometers. These methods have advantages such as fast response and high local measurement accuracy, and are suitable for acquiring unsteady parameters at fixed points or in small areas. However, in internal flow systems with complex geometric boundaries and narrow channels, invasive methods are limited by the physical size and spatial arrangement density of sensors, making it difficult to achieve high spatial resolution measurements over the entire flow domain. Furthermore, the introduction of measurement probes or sensing elements often interferes with the original flow field, alters the local flow structure, and may even induce unexpected eddy shedding and flow separation phenomena, thus affecting the accuracy and reliability of the measurement results.

[0064] In recent years, with the development of high-speed cameras and computer image processing technology, non-contact optical flow field measurement technology has rapidly advanced, with Tomo-PIV (Tomo-Particle Image Velometry) and 3D Particle Tracking Velometry (3D-PTV) being representative examples. Tomo-PIV uses multiple high-speed cameras to record two consecutive sets of laser-illuminated grayscale images of particles. A spatial grayscale volume is reconstructed from the two sets of images, and then cross-correlation is performed on the grayscale volume to obtain a three-dimensional velocity vector field. To obtain a reliable velocity vector, the particle images used in this technology must have a high particle concentration. However, in actual complex three-dimensional flows, when there are regions where particles are difficult to enter, such as boundary layers, strong shear zones, or vortex cores, the local particle concentration drops significantly, resulting in insufficient grayscale signal strength. This prevents Tomo-PIV from obtaining reliable velocity measurement data in these regions.

[0065] Three-dimensional particle tracking velocimetry (3D PPL) focuses on identifying and tracking individual tracer particles. It acquires displacement vectors and constructs a velocity field point-by-point by matching the spatial positions of particles across consecutive frames. 3D PPL offers high measurement accuracy and spatial resolution under low particle concentration conditions. However, as particle concentration increases, particle overlap and trajectory intersections intensify, significantly increasing matching ambiguity and the probability of incorrect matching, thus leading to increased velocity field reconstruction errors. Therefore, while tomographic particle image velocimetry and 3D PPL differ in their underlying principles, both are highly sensitive to the uniformity of particle spatial distribution, making it difficult to maintain accuracy across the entire flow field in non-uniform flow fields with significant gradients in particle concentration distribution.

[0066] In complex flow systems, there is a common problem of non-uniform particle distribution caused by flow separation, rotation effects, and strong acceleration regions. Therefore, it is urgent to develop high-precision and robust velocity measurement methods that can adapt to such particle distribution characteristics, so as to achieve accurate capture of complex flow structures and their evolution mechanisms across the entire domain, and provide key data support for a deeper understanding of the nature of flow, verification of numerical models, and optimization of system performance.

[0067] This invention designs a flow field measurement region segmentation method applicable to complex flow fields. It can allocate suitable measurement regions for tomographic particle image velocimetry and 3D particle tracking velocimetry, laying the foundation for obtaining accurate flow information of loaded 3D flow fields. The flow field measurement region segmentation method provided in this application relies on the Multiplicative Algebraic Reconstruction Technique (MART) algorithm to reconstruct the grayscale information of the measurement volume and obtain the coordinate information of spatial particles using a dynamic thresholding method. Then, the entire measurement region is segmented according to the spatial distribution information of the particles, and a suitable velocity measurement method is selected in different regions to improve the overall measurement accuracy. Complex 3D flow environments pose certain challenges to the accurate segmentation of measurement regions. For example, the number of particles in low-concentration regions is too small, and boundary effects and the randomness of particle distribution can degrade the boundary quality of the identification region. This invention introduces morphological processing and a concentration gradient-based identification region optimization method to improve the accuracy and boundary quality of the measurement region segmentation. Furthermore, this algorithm also introduces a concentration gradient method to further eliminate the mutual influence between the concentration values ​​of particles in different concentration regions. The flow field measurement region segmentation method provided in this application can accurately segment the measurement region, and has high recognition accuracy and strong robustness.

[0068] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will be described below with reference to the accompanying drawings.

[0069] Figure 1 This is a flowchart illustrating the flow field measurement region segmentation method provided in the embodiments of this application, as shown below. Figure 1 As shown, the method includes:

[0070] S101. Obtain particle concentration information at the spatial location of particles in the flow field.

[0071] Specifically, such as Figure 2As shown, tracer particles are first uniformly dispersed into the fluid, and then illuminated by a volumetric laser light source with a certain thickness. Images are simultaneously acquired from multiple perspectives using at least four cameras. In the image post-processing stage, based on the spatial location calibration information, a multiplicative algebraic reconstruction algorithm is used to reconstruct a three-dimensional grayscale volume. The particle coordinates are obtained from the grayscale voxels in the three-dimensional grayscale volume, and the particle concentration information at the spatial location of the particles can be further calculated.

[0072] S102. Based on the boundary concentration threshold corresponding to the flow field, the flow field is divided into regions according to the particle concentration information to obtain the first boundary between the high particle concentration region and the low particle concentration region in the flow field.

[0073] Specifically, based on the boundary concentration threshold, the flow field can be divided into regions. Regions with particle concentrations higher than the boundary concentration threshold are high particle concentration regions, and regions with particle concentrations lower than the boundary concentration threshold are low particle concentration regions. The boundary line between the high particle concentration region and the low particle concentration region is the first boundary.

[0074] For example, the first boundary is obtained through binarization. During binarization, a fixed threshold DBV = 3 / 1024H is used, where H is the number of voxels in the thickness direction of the measurement area. It is generally believed that when the particle concentration is below this value, the Tomo-PIV technique will fail to obtain a clear correlation peak due to the insufficient number of particles in the query window, resulting in significant errors in the velocity measurement results. Therefore, three-dimensional particle tracking velocimetry (3D-PTV) is required in these areas. For the concentration data after binarization using the fixed threshold, since low-concentration areas typically have a small total volume and a small number of particles, their boundaries are significantly affected by the randomness of particle distribution, resulting in poor boundary quality. Therefore, subsequent boundary processing mainly focuses on low-concentration areas.

[0075] S103. Morphological correction is performed on the first boundary to obtain the second boundary.

[0076] Specifically, morphological correction is an image processing technique based on mathematical morphology theory. It uses pre-defined structural elements to perform specific transformations on the image to correct morphological defects in segmented regions and optimize image structure. Its core is the use of basic operations such as erosion, dilation, opening, and closing to achieve effects such as noise reduction, hole filling, edge smoothing, and breaking connections. The second boundary after morphological correction is smoother and more accurate than the first boundary, thus making the segmentation of the flow field measurement region more precise.

[0077] S104. Based on the second boundary, determine the flow field measurement region segmentation result.

[0078] Specifically, based on the second boundary, the flow field measurement area is divided into a high-particle-concentration region and a low-particle-concentration region. Furthermore, tomographic particle image velocimetry is used for measurement in the high-particle-concentration region, while three-dimensional particle tracking velocimetry is used in the low-particle-concentration region, achieving accurate measurement in different measurement areas.

[0079] The flow field measurement region segmentation method provided in this application segmentes the flow field to obtain a first boundary by dividing the flow field through a boundary concentration threshold, and then further corrects the first boundary through a morphological algorithm to obtain a more accurate second boundary. The measurement region divided according to the second boundary can provide accurate measurement basis for subsequent flow field measurements, thereby improving the accuracy of flow field measurements.

[0080] In one implementation, morphological correction is performed on the first boundary to obtain the second boundary, including:

[0081] Merging adjacent high-particle-concentration regions in the flow field yields a high-particle-concentration connected region, and merging adjacent low-particle-concentration regions in the flow field yields a low-particle-concentration connected region.

[0082] In high-particle-concentration or low-particle-concentration connected regions, connected regions with a volume smaller than a volume threshold are identified as invalid connected regions.

[0083] Delete the region corresponding to the invalid connected region from the high particle concentration region or the low particle concentration region to obtain the initial corrected boundary between the high particle concentration region and the low particle concentration region.

[0084] The second boundary is obtained based on the initial corrected boundary.

[0085] Specifically, due to the randomness of particle distribution, even in high-particle-concentration regions segmented by boundary concentration threshold binarization, a small number of particles may have concentrations below the selected threshold, and these particles may be distributed arbitrarily within the high-particle-concentration region. On the other hand, particles with abnormal concentration values ​​may appear during particle concentration acquisition. These particles typically appear in small clusters, with a size not exceeding five particles. The presence of these low-concentration particles leads to incorrectly identified regions. These incorrectly identified regions do not persist in the next frame, thus preventing particle matching; furthermore, these incorrectly identified regions interfere with subsequent morphological processing, amplifying the error, and therefore need to be removed.

[0086] For example, firstly, the volume threshold is converted by the reciprocal of the boundary concentration threshold. Then, the volume of all connected components after binarization is calculated. All connected components with a volume less than 5 times the concentration threshold are deleted, and the concentration threshold is reassigned based on the concentration value of the particles in the connected components.

[0087] The flow field measurement region segmentation method provided in this application performs morphological correction based on high particle concentration regions and low particle concentration regions, discarding regions with excessively small volumes in connected domains, making the morphology of high particle concentration regions and low particle concentration regions more reasonable and accurate, and improving the accuracy of flow field measurement region segmentation.

[0088] In one implementation, obtaining the second boundary based on the initial correction boundary includes:

[0089] Morphological opening and morphological closing operations are performed on the high particle concentration region and the low particle concentration region respectively to obtain the modified boundary that morphologically modifies the first boundary.

[0090] Median filtering is applied to the corrected boundary to obtain the second boundary.

[0091] Specifically, the boundaries of the identified region obtained by binarization segmentation using a boundary concentration threshold may exhibit burrs and pits, and defects may also appear within it. These factors can cause discrepancies between the identified region and the actual high-particle-concentration or low-particle-concentration region. In the embodiments described below, the identified region is either a high-particle-concentration region or a low-particle-concentration region.

[0092] First, an opening operation is performed on the recognition region. Recognition regions obtained through binarization segmentation using a concentration threshold often exhibit defects such as burrs. Morphological opening operations can remove these erroneous structures from the recognition region while maintaining its overall position, shape, and volume. The morphological opening operation process involves performing an erosion operation followed by a dilation operation.

[0093] Then, a closing operation is performed on the recognition region. Recognition regions obtained by binarization segmentation using a concentration threshold often have defects such as pits. Morphological closing operations can fill in the pits and other erroneous structures in the recognition region while maintaining the overall position, shape, and volume of the recognition region.

[0094] Both morphological closing and morphological opening operations are based on erosion and dilation operations. The process of morphological closing operation is to perform dilation operation first and then erosion operation, while the process of morphological opening operation is to perform erosion operation first and then dilation operation.

[0095] The erosion operation proceeds as follows: traverse the entire connected component using a pre-selected structuring element; if a structuring element contains a value of 0, set all voxels within that element to 0. The dilation operation proceeds as follows: traverse the entire connected component using a pre-selected structuring element; if a structuring element contains a value of 0, set all voxels within that element to 1.

[0096] After performing morphological opening and closing operations, median filtering is required to smooth the boundaries of the recognition region. The median filtering process involves traversing the voxels throughout the connected region using a pre-selected structuring element. In this embodiment, a 5×5×5 cube is used as the structuring element. For each voxel, all values ​​in the structuring element centered on it are sorted in ascending order, and the middle value among the sorted values ​​is selected as the voxel's value. After median filtering, the boundaries of the recognition region can be further smoothed, improving boundary quality and making it more consistent with reality.

[0097] The flow field measurement region segmentation method provided in this application uses morphological opening operation to eliminate burrs on the first boundary, uses morphological closing operation to eliminate pits on the first boundary, and further performs median filtering on the morphologically corrected result to make the boundary of the identified region smoother and more accurate.

[0098] In one implementation, determining the flow field measurement region segmentation result based on the second boundary includes:

[0099] Based on the average particle concentration in the high-particle-concentration region and the low-particle-concentration region divided by the second boundary, determine the particle concentration ratio between the high-particle-concentration region and the low-particle-concentration region.

[0100] Based on the correlation between the particle concentration ratio and the boundary concentration threshold, the boundary concentration threshold associated with the particle concentration ratio is updated to the boundary concentration threshold corresponding to the flow field.

[0101] Based on the boundary concentration threshold corresponding to the flow field, the second boundary is iteratively optimized to obtain the flow field measurement region segmentation result. The iterative optimization process includes correcting and optimizing the second boundary layer by layer outward from the region divided by the second boundary, and determining whether the average particle concentration of the corrected region satisfies the relationship between the corresponding region and the boundary concentration threshold. If it satisfies, the flow field measurement region segmentation result is obtained; if it does not satisfy, the second boundary is corrected and optimized layer by layer outward from the region divided by the second boundary.

[0102] After morphological correction, the recognition region has a good boundary, but its volume is not accurate at this point. This is because at the boundary between the low-concentration and high-concentration regions, the particle concentration data on both sides of the interface influence each other during interpolation, causing the concentration data at the boundary to transition in a near-linear manner rather than abruptly. When the particle concentration ratio between the low-concentration and high-concentration regions... Concentration changes at the regional boundary under different numerical conditions. With particle concentration ratio Changes such as Figure 3 As shown. Figure 3The data in the figure were obtained through statistical calculations, with the raw data derived from a theoretical particle set generated by the program. The particle concentration at different locations was normalized using the following concentration normalization formula:

[0103]

[0104] in, The result is the normalized particle concentration at different locations. This represents the particle concentration at different locations. This represents the average particle concentration in the low particle concentration region. This represents the average particle concentration in the high particle concentration region.

[0105] The curve showing the change of the normalized concentration value at the interface with the particle concentration ratio is as follows: Figure 4 As shown. The boundary concentration threshold, limited by the applicability of the velocity measurement algorithm, is not always equal to the actual concentration value at the interface. In this case, it is necessary to optimize the recognition region based on the concentration gradient. Taking the case where the boundary concentration threshold is less than the actual interface concentration as an example, the volume of the low-particle-concentration region in the recognition region obtained through binarization will be smaller than the volume of the actual region. The recognition region optimization process is as follows:

[0106] 1) Calculate the average particle concentration in the high-particle-concentration region and the low-particle-concentration region respectively, obtain the concentration ratio between the two regions, and then use the concentration ratio to... Figure 4 The normalized concentration at the boundary is obtained from the curve, and the boundary concentration threshold is updated according to the concentration normalization formula above.

[0107] 2) Expand the low particle concentration region in the existing identification area outward layer by layer, and determine whether the particle concentration value in the spatial grid point corresponding to these expanded voxels is less than the boundary concentration threshold. If the particle concentration value is less than the boundary concentration threshold, then include these voxels in the low particle concentration region.

[0108] 3) Recalculate the average concentration inside and outside the identified area, and update the boundary concentration threshold;

[0109] 4) Repeat steps 2) to 3) until the volume change of the identified region before and after one iteration is less than the convergence criterion.

[0110] If the boundary concentration threshold is greater than the actual interface concentration, the expansion in step 2) is changed to shrinking a voxel inward, and it is determined whether the concentration value corresponding to these shrunken voxels is greater than the actual concentration threshold. If the concentration value is greater than the concentration threshold, these voxels are excluded from the low particle concentration region.

[0111] The flow field measurement region segmentation method provided in this application, after being optimized by a concentration gradient-based identification region method, allows the boundary of the identification region to better approximate the actual concentration boundary, greatly improving the accuracy of the identification region.

[0112] In one implementation, obtaining particle concentration information at the spatial location of particles in the flow field includes:

[0113] Image calibration establishes a mapping relationship between the spatial location of the particle and the pixel space of the grayscale image, thereby obtaining the grayscale image corresponding to the spatial location of the particle.

[0114] Based on multiplicative algebraic reconstruction technology, a gray volume corresponding to a spatial location is constructed from a grayscale image;

[0115] In a grayscale volume, adjacent voxels with similar grayscale values ​​are merged to obtain the particles corresponding to the voxels.

[0116] Based on the particles, obtain the particle concentration information of the spatial location of the particles in the flow field.

[0117] Specifically, image calibration is performed to establish the relationship between physical space and the pixel space of the grayscale image. The most crucial aspect of this process is obtaining information about the thickness direction of the measurement area. The thickness direction of the measurement area is defined as the z-direction, perpendicular to the calibration plate plane, and the calibration plane is defined as the xy-plane. During calibration, the target plate is moved at equal intervals along the z-direction to traverse the entire thickness direction of the measurement space. At each calibration position, the dot matrix information of the calibration plate is identified, and a third-order precision polynomial is used to approximate the mapping function between the three-dimensional spatial coordinates and the two-dimensional pixel coordinates, as follows:

[0118]

[0119]

[0120] Where (X,Y) are two-dimensional pixel coordinates, and (x,y) are coordinates in the xy plane at a certain z value, with coefficients... and These are calibration coefficients. Given a sufficient number of calibration points, high-precision coefficients can be obtained by solving the system of equations formed by the two formulas above using least-squares fitting. and This establishes the relationship between physical space and pixel space.

[0121] In addition, internal parameters such as focal length, principal point, and distortion of each camera, as well as external parameters such as the relative position and attitude between cameras, can be used to establish an accurate mapping relationship between two-dimensional pixel coordinates and three-dimensional world coordinates.

[0122] Furthermore, multiplicative algebraic reconstruction techniques are used to reconstruct the grayscale volume. In multiplicative algebraic reconstruction, the grayscale value of each pixel in the projected image (i.e., the captured particle image) is considered to be a weighted integral of the grayscale values ​​of the voxels in the grayscale volume along the line of sight (LOS). Mathematically, the projection equation can be expressed as:

[0123]

[0124] In the formula, I represents the image gray level, and E represents the gray level in the space. This refers to the weights of specific voxels relative to pixels. Under actual experimental conditions, the number of unknowns in the system of projection equations far exceeds the number of equations themselves, classifying it as an indeterminate problem requiring iterative solutions. The specific iterative process is as follows:

[0125] i. Begin the k-th iteration:

[0126] ii. Integrate along the line of sight for each camera m (where i is the pixel index):

[0127]

[0128] iii. Update the value of each voxel j:

[0129]

[0130] End loop iii,

[0131] End loop ii,

[0132] End loop i.

[0133] in, The relaxation factor is typically set to 1. This invention uses a Gaussian function as the point propagation function for summation. After several iterations, the grayscale volume can be reconstructed.

[0134] Then, the spatial coordinates of each particle are extracted from the grayscale volume. This process involves the following steps:

[0135] (1) Identify particle boundaries using a dynamic thresholding method. Find all local maxima in the entire grayscale volume, i.e., the peak points of brightness. Then, expand the boundary of each peak voxel outward by one voxel size and check each new voxel. If the ratio of the voxel grayscale in the boundary to the peak voxel grayscale is greater than a preset contrast threshold, then the voxel and the peak voxel point are considered to belong to the same particle; otherwise, the voxel is considered not to belong to the particle. The implementation process is as follows: Figure 5 As shown:

[0136] (2) Calculate the gray centroid of all voxels in each particle. The calculation formula is as follows:

[0137]

[0138]

[0139]

[0140] in For each particle's gray-level centroid coordinates, The coordinates of a voxel in each particle. Let be the total volume of the particle. The gray-scale centroid coordinates of the particle, calculated from this, are considered to be the particle's spatial coordinates.

[0141] Based on the spatial coordinates of the particles, the particle concentration information within the region can be quantified.

[0142] The flow field measurement region segmentation method provided in this application realizes the mapping between spatial location and grayscale pixels through image calibration, providing a precise coordinate reference for subsequent 3D reconstruction; it constructs grayscale volumes based on multiplicative algebraic reconstruction technology, transforming 2D image information into 3D spatial voxel data, filling the information gap between 2D observation and 3D space; it merges adjacent voxels with similar grayscale to identify particles, effectively eliminating noise interference, accurately defining particle boundaries, and ensuring the accuracy of particle identification; finally, it obtains the concentration information of the flow field spatial location based on the identified particles, realizing a closed-loop process from image acquisition, 3D reconstruction to concentration quantification, providing high-precision and traceable quantitative data support for flow field characteristic analysis, while improving the reliability of particle concentration detection results.

[0143] In one implementation, obtaining particle concentration information at the spatial location of particles in the flow field includes:

[0144] Establish the Vino polyhedron corresponding to the particle;

[0145] The local particle concentration is determined by the reciprocal of the volume of the Vino polyhedron;

[0146] The local particle concentration is interpolated to spatial grid points corresponding to the particle physical coordinates using linear interpolation.

[0147] The local particle concentration at the spatial grid points is smoothed to obtain the particle concentration information at the spatial location of the particles.

[0148] Specifically, to generate the Vino polyhedra corresponding to all particles, firstly, a Delaunay tetrahedron network is generated. The Delaunay tetrahedron network is composed of many tetrahedra, which are generated based on the following criteria: (1) each vertex of the tetrahedron is a point in the point set; (2) there are no other points within the circumsphere of any tetrahedron; (3) in the tetrahedron network, the minimum angle of the triangle in the Delaunay triangulation network is the largest. Based on the Delaunay tetrahedron network, the center of the circumsphere of each tetrahedron is calculated. These centers are the vertices of the Vino polyhedra. Finding the centers of the circumspheres of all Delaunay tetrahedra with a certain point in the point set as the vertex is the result of finding all the vertices of the Vino polyhedra corresponding to that point.

[0149] by Figure 6 For example, Figure 6 The dots in the left image represent particles, such as... Figure 6 As shown in the middle diagram, three adjacent particles are connected to form a triangular mesh. The position of the circumcenter of each triangle in the mesh is determined, and then... Figure 6 As shown in the right figure, the Vinno polyhedron is generated by using the circumcenters of all the triangles as vertices.

[0150] The local particle concentration is characterized by the reciprocal of the Vino polygon volume of a single tracer particle, defined as the Voronoi density (DBV) concentration, and its formula is:

[0151]

[0152] Among them, S(p i Let be the volume of the i-th Vino polyhedron, in voxels. -1 .

[0153] Taking into account both interpolation accuracy and computational complexity, linear interpolation is used to interpolate the local particle concentration of all particles to spatial grid points corresponding to voxel physical coordinates. This step can eliminate interference that is difficult to eliminate through morphological processing due to the randomness of particle distribution and mutual interference between particles at different concentration boundaries.

[0154] The concentration at the grid points is then smoothed. Due to the randomness of particle distribution in space, the data at the grid points typically exhibits significant fluctuations, which is detrimental to the subsequent segmentation of the computational region. This can easily lead to misidentification of high-concentration regions as low-concentration regions and degrade the quality of the identified region boundaries. In this embodiment, Gaussian filtering is applied to the concentration data in the grid. Gaussian filtering is a linear smoothing filter that uses a function following a Gaussian distribution as its convolution kernel. For three-dimensional space, the expression for the Gaussian function is:

[0155]

[0156] in Let be the spatial coordinates, and σ be the standard deviation of the Gaussian distribution, which is defaulted to 0.65 in this embodiment. A 5×5×5 window is selected as the convolution kernel during the smoothing process. The smoothed value at the center point of the window is obtained by multiplying the original data in the window with the Gaussian weights in the 3D Gaussian kernel and then summing the results.

[0157] The flow field measurement region segmentation method provided in this application transforms the discrete spatial distribution of particles into a unit division with clear geometric boundaries by establishing a Vino polyhedron corresponding to each particle. This provides an intuitive and non-overlapping spatial carrier for local concentration calculation. Defining the local particle concentration as the reciprocal of the Vino polyhedron volume directly reflects the density of the particle spatial arrangement, avoiding the subjectivity of manually setting the statistical radius in traditional methods. Linear interpolation maps discrete local concentration values ​​to regular spatial grid points, realizing the conversion from discrete data to a continuous field and meeting the flow field analysis's requirement for gridded data. Finally, smoothing the grid point concentration effectively reduces local fluctuation noise caused by random particle distribution, improving the continuity, stability, and interpretability of the concentration field data, and providing a high-precision and highly reliable quantitative basis for subsequent flow field measurements.

[0158] In one implementation, establishing the Vino polyhedron corresponding to the particle includes:

[0159] The particle is translated along a preset direction to obtain the translated particle.

[0160] Centered on the particle, construct an extended spatial point set based on the particle and its translation;

[0161] Generate the Vino polyhedron for the extended space point set and establish the Vino polyhedron corresponding to the particle.

[0162] Ideally, a Vino polyhedron should extend infinitely in space, but in reality, it can only be generated within a finite space. In this case, Vino polyhedra near the boundary will extend infinitely; this phenomenon is called the boundary effect of Vino polyhedra. Simply truncating the Vino polyhedra at the boundary will cause a significant difference between the shape of the polyhedron at the boundary and the shape of the polyhedra inside. In this embodiment, the original particle point set is translated along the three coordinate axes to construct an extended spatial point set structure centered on the original point set and arranged in a 3×3×3 pattern. Then, Vino polyhedra are generated. This operation effectively eliminates the influence of the boundary effect, ensuring that all Vino polyhedra have good shapes. Vino polyhedra generated through this process are as follows: Figure 7 As shown.

[0163] The flow field measurement region segmentation method provided in this application translates the particle point set to construct an extended spatial point set, and generates a Vinno polyhedron based on the extended spatial point set. This can effectively reduce the boundary effect of the Vinno polyhedron and improve the accuracy of particle concentration measurement.

[0164] To further demonstrate the practical effects of the present invention, this application provides a specific example, using the flow field measurement region segmentation method provided in the above embodiments to segment the measurement region in a synthetic measurement space of 1024*1024*200 pixels. For example... Figure 8 As shown, a spherical low-particle-concentration region with a radius of 75 pixels exists within the measurement area. The particle concentration in this low-concentration region is 1 × 10⁻⁶. -5 voxel -1 In the high-concentration region, the particle concentration is 5 × 10⁻⁶. -5 voxel -1 The case of Vino polyhedra generated based on particle point sets in low-concentration regions is as follows: Figure 9 As shown. The recognition region obtained by binarizing the concentration data using a fixed threshold, and the recognition region after morphological processing and median filtering are shown in the figure. Figure 10 As shown. For the recognition region, its recognition quality can be evaluated by calculating the intersection-union ratio (IUU) between the recognition region and the theoretical region, using the following formula:

[0165]

[0166] Regarding the Intersection over Union (IoU) value, it is generally believed that when IoU ≥ 0.8, the segmented recognition region highly overlaps with the theoretical region, the boundary positioning is accurate, and it can fully meet the requirements of subsequent PIV / PTV fusion measurement for region segmentation accuracy.

[0167] The cross-union ratio (CUC) between the identified region and the theoretical region obtained by binarizing the concentration data with a fixed threshold is only 0.433, which is a very low value, indicating that the obtained identified region cannot meet the needs of practical use.

[0168] Further optimization of the recognition region based on concentration gradient and its comparison with the theoretical region are shown below. Figure 11 As shown in Figures (a) and (b).

[0169] The intersection-union ratio (IUGR) between the identified region and the theoretical region obtained by the flow field measurement region segmentation method provided in this application can reach 0.893, indicating a high degree of overlap between the identified region and the theoretical region, thus demonstrating the accuracy of the identification algorithm. Furthermore, statistical analysis of the low-concentration particles obtained from the identified region and the theoretical low-concentration particles yields an identification accuracy of 0.842, further corroborating the accuracy of the algorithm.

[0170] Figure 12 This is a schematic diagram of the flow field measurement region segmentation device provided in the embodiments of this application, as shown below. Figure 12 As shown, the flow field measurement region segmentation device 120 provided in this embodiment includes:

[0171] The acquisition module 1201 is used to acquire particle concentration information at the spatial location of particles in the flow field;

[0172] The segmentation module 1202 is used to segment the flow field based on the boundary concentration threshold corresponding to the flow field and according to the particle concentration information to obtain the first boundary between the high particle concentration region and the low particle concentration region in the flow field.

[0173] The correction module 1203 is used to perform morphological correction on the first boundary to obtain the second boundary;

[0174] The determination module 1204 is used to determine the flow field measurement region segmentation result based on the second boundary.

[0175] In one implementation, the correction module 1203 is specifically used for:

[0176] Merging adjacent high-particle-concentration regions in the flow field yields a high-particle-concentration connected region, and merging adjacent low-particle-concentration regions in the flow field yields a low-particle-concentration connected region.

[0177] In high-particle-concentration or low-particle-concentration connected regions, connected regions with a volume smaller than a volume threshold are identified as invalid connected regions.

[0178] Delete the region corresponding to the invalid connected region from the high particle concentration region or the low particle concentration region to obtain the initial corrected boundary between the high particle concentration region and the low particle concentration region.

[0179] The second boundary is obtained based on the initial corrected boundary.

[0180] In one implementation, the correction module 1203 is further configured to:

[0181] Morphological opening and morphological closing operations are performed on the high particle concentration region and the low particle concentration region respectively to obtain the modified boundary that morphologically modifies the first boundary.

[0182] Median filtering is applied to the corrected boundary to obtain the second boundary.

[0183] In one implementation, the determining module 1204 is specifically used for:

[0184] Based on the average particle concentration in the high-particle-concentration region and the low-particle-concentration region divided by the second boundary, determine the particle concentration ratio between the high-particle-concentration region and the low-particle-concentration region.

[0185] Based on the correlation between the particle concentration ratio and the boundary concentration threshold, the boundary concentration threshold associated with the particle concentration ratio is updated to the boundary concentration threshold corresponding to the flow field.

[0186] Based on the boundary concentration threshold corresponding to the flow field, the second boundary is iteratively optimized to obtain the flow field measurement region segmentation result. The iterative optimization process includes correcting and optimizing the second boundary layer by layer outward from the region divided by the second boundary, and determining whether the average particle concentration of the corrected region satisfies the relationship between the corresponding region and the boundary concentration threshold. If it satisfies, the flow field measurement region segmentation result is obtained; if it does not satisfy, the second boundary is corrected and optimized layer by layer outward from the region divided by the second boundary.

[0187] In one implementation, the acquisition module 1201 is specifically used for:

[0188] Image calibration establishes a mapping relationship between the spatial location of the particle and the pixel space of the grayscale image, thereby obtaining the grayscale image corresponding to the spatial location of the particle.

[0189] Based on multiplicative algebraic reconstruction technology, a gray volume corresponding to a spatial location is constructed from a grayscale image;

[0190] In a grayscale volume, adjacent voxels with similar grayscale values ​​are merged to obtain the particles corresponding to the voxels.

[0191] Based on the particles, obtain the particle concentration information of the spatial location of the particles in the flow field.

[0192] In one embodiment, the acquisition module 1201 is further configured to:

[0193] Establish the Vino polyhedron corresponding to the particle;

[0194] The local particle concentration is determined by the reciprocal of the volume of the Vino polyhedron;

[0195] The local particle concentration is interpolated to spatial grid points corresponding to the particle physical coordinates using linear interpolation.

[0196] The local particle concentration at the spatial grid points is smoothed to obtain the particle concentration information at the spatial location of the particles.

[0197] In one embodiment, the acquisition module 1201 is further configured to:

[0198] The particle is translated along a preset direction to obtain the translated particle.

[0199] Centered on the particle, construct an extended spatial point set based on the particle and its translation;

[0200] Generate the Vino polyhedron for the extended space point set and establish the Vino polyhedron corresponding to the particle.

[0201] The flow field measurement region segmentation device provided in this embodiment can execute the method provided in the above method embodiment. Its implementation principle and technical effect are similar, and will not be described in detail here.

[0202] Figure 13 This is a schematic diagram of the structure of the flow field measurement region segmentation device provided in an embodiment of this application. Figure 13 As shown, the flow field measurement region segmentation device 130 provided in this embodiment includes at least one processor 1301 and a memory 1302. Optionally, the flow field measurement region segmentation device 130 further includes a communication interface 1303. The processor 1301, memory 1302, and communication interface 1303 are connected via a communication bus 1304.

[0203] In a specific implementation, at least one processor 1301 executes computer execution instructions stored in memory 1302, causing at least one processor 1301 to perform the above-described method.

[0204] The specific implementation process of processor 1301 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0205] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0206] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0207] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0208] This application also provides a computer program product, including a computer program that, when executed, implements the above-described method.

[0209] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed, implement the above-described method.

[0210] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0211] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an application-specific integrated circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.

[0212] The division of units is merely a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0213] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0214] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0215] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0216] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0217] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. A method for segmenting a flow field measurement region, characterized in that, include: Obtain particle concentration information based on the spatial location of particles in the flow field; Based on the boundary concentration threshold corresponding to the flow field, the flow field is divided into regions according to the particle concentration information to obtain the first boundary between the high particle concentration region and the low particle concentration region in the flow field. The first boundary is morphologically corrected to obtain the second boundary; Based on the second boundary, the flow field measurement region segmentation result is determined.

2. The flow field measurement region segmentation method according to claim 1, characterized in that, The step of performing morphological correction on the first boundary to obtain the second boundary includes: The adjacent high-particle-concentration regions in the flow field are merged to obtain a high-particle-concentration connected region, and the adjacent low-particle-concentration regions in the flow field are merged to obtain a low-particle-concentration connected region. In the high-particle-concentration or low-particle-concentration connected regions, connected regions with a volume smaller than a volume threshold are identified as invalid connected regions. Delete the region corresponding to the invalid connected component from the high particle concentration region or the low particle concentration region to obtain the initial corrected boundary between the high particle concentration region and the low particle region. The second boundary is obtained based on the initial corrected boundary.

3. The flow field measurement region segmentation method according to claim 2, characterized in that, Obtaining the second boundary based on the initial corrected boundary includes: Morphological opening and morphological closing operations are performed on the high particle concentration region and the low particle concentration region respectively to obtain a modified boundary that morphologically modifies the first boundary. The corrected boundary is then subjected to median filtering to obtain the second boundary.

4. The flow field measurement region segmentation method according to any one of claims 1 to 3, characterized in that, The step of determining the flow field measurement region segmentation result based on the second boundary includes: Based on the average particle concentration in the high-particle-concentration region and the low-particle-concentration region divided by the second boundary, determine the particle concentration ratio between the high-particle-concentration region and the low-particle-concentration region. Based on the correlation between the particle concentration ratio and the boundary concentration threshold, the boundary concentration threshold associated with the particle concentration ratio is updated to the boundary concentration threshold corresponding to the flow field. Based on the boundary concentration threshold corresponding to the flow field, the second boundary is iteratively optimized to obtain the flow field measurement region segmentation result. The iterative optimization process includes correcting and optimizing the second boundary layer by layer outward from the region divided according to the second boundary, and determining whether the average particle concentration of the corrected region satisfies the relationship between the corresponding region and the boundary concentration threshold. If it satisfies the relationship, the flow field measurement region segmentation result is obtained. If it does not satisfy the relationship, the step of correcting and optimizing the second boundary layer by layer outward from the region divided according to the second boundary is performed.

5. The flow field measurement region segmentation method according to any one of claims 1 to 3, characterized in that, The process of obtaining particle concentration information based on the spatial location of particles in the flow field includes: Image calibration establishes a mapping relationship between the spatial location of the particle and the pixel space of the grayscale image, thereby obtaining the grayscale image corresponding to the spatial location of the particle. Based on the multiplicative algebra reconstruction technique, a gray volume corresponding to the spatial location is constructed according to the gray image; The voxels contained in the gray volume that are adjacent and have similar gray values ​​are merged to obtain the particles corresponding to the voxels. Based on the particles, obtain the particle concentration information of the spatial location of the particles in the flow field.

6. The flow field measurement region segmentation method according to claim 5, characterized in that, The process of obtaining particle concentration information based on the spatial location of particles in the flow field includes: Establish the Vino polyhedron corresponding to the particle; The local particle concentration is determined based on the reciprocal of the volume of the Vino polyhedron; The local particle concentration is interpolated to spatial grid points corresponding to the particle physical coordinates using linear interpolation. The local particle concentration at the spatial grid point is smoothed to obtain the particle concentration information at the spatial location of the particle.

7. The flow field measurement region segmentation method according to claim 6, characterized in that, Construct the Vino polyhedron corresponding to the particle, including: The particle is translated along a preset direction to obtain the translated particle. Centered on the particle, construct an extended spatial point set based on the particle and the translated particle; Generate a Vinno polyhedron for the extended space point set and establish the Vinno polyhedron corresponding to the particle.

8. A flow field measurement region segmentation device, characterized in that, include: Memory, processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the method as described in any one of claims 1-7.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed, are used to implement the method as described in any one of claims 1-7.

10. A computer program product, characterized in that, Includes a computer program, which, when executed, implements the method according to any one of claims 1-7.

Images