A device and method for measuring the trajectory and attitude of irregular sand particles

Abstract

The application discloses a device and method for measuring the motion trajectory and posture of irregular sand particles, and belongs to the field of measuring the motion trajectory of irregular particles. The device comprises an airflow generating device, a connecting pipe, a sand particle spreading device, an airflow pipe, a laser light source, a shaping element, a camera and a data processing unit. The airflow generating device is connected with the airflow pipe through the connecting pipe. The airflow pipe is sequentially divided into a fully developed zone, an impact observation zone and a discharge zone along the airflow flow direction. The sand particle spreading device is arranged at the starting position of the fully developed zone. The impact observation zone is provided with a blade. The laser light source is combined with the shaping element to generate a laser beam and shape the laser beam. The camera is arranged outside the airflow pipe. The data processing unit is used for quantitatively measuring the motion trajectory and posture change process of the irregular sand particles according to a speckle image sequence. The application can provide reliable experimental data support for the research on the sand particle impact and erosion mechanism.

Description

Technical Field

[0001] This invention relates to the field of measuring the motion trajectory of irregular particles, and particularly to a device and method for measuring the motion trajectory and attitude of irregular sand particles. Background Technology

[0002] With the increasing demand for aero-engines operating in environments with sand particles, high temperatures, and complex airflow, the erosion and damage caused by sand particles to internal engine components, especially compressor and turbine blades, is becoming increasingly prominent. Sand particles, carried by high-speed airflow, impact the blade surface, triggering a series of complex physical processes such as material removal, surface roughening, and coating failure, severely affecting the engine's aerodynamic performance, structural strength, and service life. Therefore, a thorough understanding of the impact and erosion mechanisms between sand particles and blades is a crucial research area in the field of aero-engine protection design and life assessment.

[0003] Existing research generally agrees that the erosion process of sand particles on blades is not only closely related to the particle velocity, size, and material properties of the sand particles, but also highly dependent on the spatial trajectory, attitude, and incident angle of the sand particles at the moment of impact. Especially for irregularly shaped sand particles commonly found in natural sand environments, their flight is often accompanied by complex tumbling, rotation, and attitude evolution behaviors. These factors directly affect the stress distribution, energy dissipation mode, and subsequent fragmentation, rebound, or secondary splashing behavior in the contact area between the sand particles and the blade. However, these key physical quantities remain difficult to accurately obtain in current experimental studies.

[0004] Currently, experimental methods used to study the mechanism of sand erosion mainly include high-speed photography, single-view or dual-view imaging, erosion test benches, and post-impact surface morphology analysis. These techniques can usually obtain the macroscopic trajectory or average impact velocity of sand particles, but they are limited by factors such as imaging angle, depth of field, temporal resolution, and sand particle size, making it difficult to simultaneously analyze the three-dimensional attitude changes, instantaneous incident angle, and spin characteristics of sand particles before and after impacting the blade. In addition, most existing methods tend to simplify sand particles into spherical or axisymmetric particles during the modeling process, ignoring the influence of irregular geometry on attitude evolution and impact behavior, thus leading to significant deviations between experimental results and real-world conditions.

[0005] Furthermore, after a high-speed collision between sand grains and blades, the sand grains may exhibit complex phenomena such as breakage, peeling, rebound, or multi-body splitting. The formation mechanism of these phenomena is closely related to their attitude state and contact mode before and after the collision. However, due to the lack of continuous and quantitative measurement methods for the trajectory and attitude of sand grains, existing technologies struggle to establish a correspondence between the pre-collision motion state and the post-collision evolution behavior, thus limiting a deeper understanding of the sand grain erosion mechanism.

[0006] Therefore, there is an urgent need for a method and device that can accurately measure the motion trajectory and attitude information of irregular sand particles in complex aerodynamic environments, so as to make up for the shortcomings of existing experimental techniques in attitude analysis and impact behavior characterization, and provide more comprehensive and reliable experimental data support for the study of sand particle erosion mechanism in aero-engines. Summary of the Invention

[0007] The purpose of this invention is to provide a device and method for measuring the motion trajectory and attitude of irregular sand particles, aiming to solve the problem that it is difficult to obtain key information such as the motion trajectory, instantaneous attitude and attitude change of sand particles before and after impacting aero-engine blades in existing sand erosion experiments, and to provide reliable experimental data support for the study of sand particle impact and erosion mechanism.

[0008] To achieve the above objectives, the present invention adopts the following technical solution: A device for measuring the motion trajectory and attitude of irregular sand particles includes an airflow generating device, a connecting pipe, a sand particle spreading device, an airflow duct, a laser light source, a shaping element, a camera, and a data processing unit. The airflow generating device is connected to the airflow pipe via a connecting pipe. The airflow pipe is divided into a fully developed zone, an impact observation zone, and a discharge zone along the airflow direction. The sand grain spreading device is set at the starting position of the fully developed zone to introduce sand grains into the airflow pipe and make the sand grains move along the direction of the airflow pipe under the action of the airflow. The impact observation zone is equipped with blades to collide with the sand grains moving with the airflow. The discharge zone is used to discharge the impacted sand grains and the debris generated by them out of the airflow pipe. The laser source and shaping element are combined to generate a laser beam and shape the laser beam so that the laser beam forms a laser illumination area covering the impact observation area, thereby illuminating the sand particles entering the laser illumination area. The camera is set outside the airflow duct and corresponds to the impact observation area. It is used to collect the scattered light signal generated by the sand particles in the laser illumination area, thereby continuously capturing the speckle image of the sand particles and forming a speckle image sequence. The data processing unit is used to obtain the motion trajectory of each sand grain during the experiment based on the speckle image sequence, obtain the two-dimensional morphology image sequence of each sand grain at different times based on the motion trajectory of the sand grain, determine the instantaneous attitude of the sand grain relative to the fixed angle of the camera at each time based on the two-dimensional morphology image sequence, and obtain the rotation amount and centroid displacement of the sand grain in three spatial directions within adjacent shooting intervals based on the instantaneous attitude, thereby realizing the quantitative measurement of the motion trajectory and attitude change process of irregular sand grains.

[0009] Furthermore, the airflow duct is a hollow structure with a square cross-section, and the airflow duct is made of transparent material.

[0010] Furthermore, the impact observation area is equipped with a blade fixing device, and the blade is installed in the impact observation area through the blade fixing device.

[0011] Furthermore, the sand spreading device is a cylindrical structure with openings at both ends. One end of the cylindrical structure is connected to the starting end of the fully developed zone of the airflow pipe, and the other end is used to load sand particles. The sand spreading device is equipped with a switching mechanism for controlling the scattering of sand particles at the end connected to the airflow pipe.

[0012] Furthermore, obtaining the trajectory of each sand grain during the experiment based on the speckle image sequence specifically involves: The speckle image sequence is processed, and each sand grain entering the impact observation area is identified by the target detection and tracking algorithm. The spatial position change information of each sand grain in the speckle image sequence is extracted, thereby obtaining the motion trajectory of each sand grain during the experiment. The step of obtaining a sequence of two-dimensional morphological images of each sand grain at different times based on the movement trajectory of the sand grains is specifically as follows: Based on the movement trajectory of the sand grains, the corresponding sand grain regions are automatically cropped in each frame of the speckle image to obtain a speckle image sequence of each sand grain at different times. The speckle image sequence at different times is then input frame by frame into a pre-trained generative network model. The generative network model is built based on a deep generative learning method and can invert the two-dimensional morphology image of the sand grain from the camera's perspective from the speckle image, thereby forming a two-dimensional morphology image sequence of each sand grain at different times.

[0013] Furthermore, determining the instantaneous attitude of the sand grain relative to a fixed viewpoint of the camera at each moment based on the two-dimensional topographic image sequence specifically involves: Based on the two-dimensional topography image sequence, a multi-view three-dimensional reconstruction method is used to perform three-dimensional topography inversion on sand grains to obtain the three-dimensional topography model of each sand grain. Based on the three-dimensional topography model of the sand grain and its corresponding two-dimensional topography image at each time, the instantaneous attitude of the sand grain relative to the fixed viewpoint of the camera at each time is determined. The process of obtaining the rotation of sand grains in three spatial directions and the displacement of their center of mass within adjacent shooting intervals based on instantaneous attitude is specifically as follows: By performing difference or differentiation on the instantaneous attitude of the same sand grain in a time series, the rotation of the sand grain in three spatial directions and the displacement of its center of mass within adjacent shooting intervals can be obtained.

[0014] A method for measuring the motion trajectory and attitude of irregular sand particles, comprising: Start the airflow generation device and introduce airflow into the airflow duct through the connecting pipe. The airflow flows sequentially through the fully developed zone, the impact observation zone, and the discharge zone along the airflow direction, forming a stable flow field inside the airflow duct. The sand particles are dropped into the airflow duct by the sand particle spreading device, and a sand particle group is formed under the combined action of gravity and airflow, and acceleration is completed in the fully developed zone; When the sand grains enter the impact observation area, the laser light source forms a laser illumination area under the action of the shaping element, irradiating the sand grains. Under synchronous triggering conditions, the camera continuously acquires speckle images of the sand grains at different times. Based on the speckle images collected at different times, the sand grain speckles in the speckle images are identified and matched. By associating the centroid positions of the sand grain speckles at consecutive times, the motion trajectory of each sand grain before and after the impact is obtained. Based on the motion trajectory, the speckle images of the same sand grain at different times are cropped and collected, and input into a pre-trained speckle to two-dimensional morphology image mapping model to obtain two-dimensional morphology images of the sand grain in different postures; based on the two-dimensional morphology images of the same sand grain in different postures, a three-dimensional morphology model of the sand grain is reconstructed through a multi-view three-dimensional reconstruction method. The three-dimensional morphology model of the sand grain is projected and matched with the two-dimensional morphology images at different times to solve the instantaneous attitude of the sand grain at each time. By the change of the instantaneous attitude between adjacent times, the rotation of the sand grain in three directions in space and the displacement of the center of mass are obtained, thus realizing the complete measurement of the motion trajectory and attitude of irregular sand grains.

[0015] Furthermore, based on the collected speckle images at different times, the sand grain speckles in the speckle images are identified and matched. By associating the centroid positions of the sand grain speckles at consecutive times, the trajectory of each sand grain before and after the impact is obtained, specifically: Preprocess the speckle images at different times; In the speckle image at the first moment after preprocessing, all sand grain speckles are identified and the centroid position of each sand grain speckle is calculated. A square search window is established with the centroid position as the center for speckle matching in subsequent time frames. In the speckle images at the second and subsequent time points, the square search window of the previous time point is used to search for sand grain speckles in the corresponding region and calculate the centroid position. At the same time, the position and range of the square search window are updated. This process is repeated until all time frames are traversed, so as to obtain the centroid position of each sand grain speckle at different time points. Connecting the centroids of the same sand grain speckle in consecutive time frames in chronological order forms the trajectory of the sand grain before and after the impact.

[0016] Further, based on the motion trajectory, the speckle images of the same sand grain at different times are cropped and collected, and input into a pre-trained speckle-to-two-dimensional morphology image mapping model to obtain two-dimensional morphology images of the sand grain in different postures; based on the two-dimensional morphology images of the same sand grain in different postures, a three-dimensional morphology model of the sand grain is reconstructed through a multi-view three-dimensional reconstruction method, specifically including: Based on the movement trajectory of the sand grains, the speckle images of the same sand grain at different times are classified and collected to form a speckle image sequence of a single sand grain. Each frame of the speckle image sequence of a single sand grain is a multi-view speckle image obtained from different equivalent viewpoints. Multi-view speckle images are input into a pre-trained speckle-to-2D topography image mapping model to generate corresponding 2D topography images of sand grains under different views. The 2D contour mask of each view is obtained through edge detection processing. By utilizing the imaging geometry of the camera, the two-dimensional contour mask is back-projected into three-dimensional space to generate the view frustum corresponding to each viewpoint. By performing spatial Boolean intersection operation on all view frustums, the three-dimensional envelope of the sand grain is obtained. The three-dimensional envelope is discretized to generate a three-dimensional voxel model, and the three-dimensional morphology model of the sand grains is obtained by smoothing filtering and void filling.

[0017] Furthermore, the process involves projecting and matching the three-dimensional morphology model of the sand grains with two-dimensional morphology images at different times to solve for the instantaneous attitude of the sand grains at each time. By analyzing the changes in the instantaneous attitude between adjacent times, the rotation of the sand grains in three spatial directions and the displacement of their center of mass are obtained, thereby achieving a complete measurement of the trajectory and attitude of irregular sand grains. Specifically: Using the two-dimensional topography image corresponding to the speckle image at the first moment as the target image, a rotation transformation and a first translation vector t0 are applied to the three-dimensional topography model of the sand grains. The rotation transformation is expressed as a revolution around... x , y , z The rotation combination of the three axes, the rotation transformation is represented by the first rotation matrix R0, so that its positive projection in the camera coordinate system is consistent with the two-dimensional shape image, then the rotation state corresponding to the rotation transformation is determined as the reference posture of the sand grain, the major axis vector and minor axis vector of the sand grain are extracted, and the reference coordinate system is constructed. Using the two-dimensional topography image corresponding to the speckle image at the second time step as the new target image, a second rotation matrix R1 and a second translation vector t1 are applied to the three-dimensional topography model of the same sand grain to make the two-dimensional projection consistent with the new target image. The major axis direction vector and minor axis direction vector after rotation are extracted to construct the current attitude coordinate system. The transformation relationship between the current attitude coordinate system and the reference coordinate system is represented as a rigid body pose transformation, including a second rotation transformation and a translation transformation. The second rotation transformation is performed using a third rotation matrix R1.01 =R1·R0 - ¹ indicates that the translation transformation uses the centroid position vector difference Δt=t1. t0 represents; The third rotation matrix R 01 Decomposed into winding x , y , z The rotation of the three coordinate axes is denoted as (Δ). θ x , Δ θ y , Δ θ z This method describes the flipping, rolling, and deflection behavior of sand grains between adjacent moments. By solving the attitude of each frame at each moment in a continuous time period and calculating the attitude change between adjacent moments, the rotation of the sand grains in three spatial directions and the displacement of the center of mass are obtained, thus realizing the complete measurement of motion trajectory and attitude evolution.

[0018] Compared with the prior art, the present invention has the following beneficial technical effects: (1) The device of the present invention has strong versatility and can be easily integrated with existing wind tunnels, erosion test benches and high-speed airflow systems without requiring major modifications to the original test system. It has good engineering applicability and promotion value. When the present invention is used, it can realize the synchronous measurement of the motion trajectory and attitude of irregular sand particles. By combining high-speed imaging, trajectory extraction and morphology inversion, the motion trajectory, instantaneous attitude and attitude change of sand particles can be obtained simultaneously within the same test framework, overcoming the limitation of existing technologies that can only obtain speed or trajectory but cannot analyze attitude information.

[0019] (2) This invention utilizes the flipping and rotation of irregular sand grains during the movement process to regard the two-dimensional shape acquired by the same camera at different times as equivalent multi-view information, thereby realizing three-dimensional shape reconstruction and attitude solution, avoiding complex multi-camera synchronization and calibration, and improving stability and engineering feasibility.

[0020] (3) The present invention is based on speckle imaging and generation network inversion of two-dimensional morphology. It does not rely on idealized assumptions about the geometry of sand grains and can be applied to irregular sand grains that are common in natural sand grains, which significantly improves the consistency between measurement results and real working conditions.

[0021] (4) This invention can not only obtain the incident angle and attitude state of sand particles at the moment of impact, but also obtain the attitude change of sand particles before and after impact (i.e. the amount of rotation of sand particles in three directions in space) and the displacement of the center of mass through time series analysis, providing direct experimental basis for revealing the physical mechanisms of sand particle breakage, rebound and secondary splashing. Attached Figure Description

[0022] The accompanying drawings are provided to further understand the invention and constitute a part of this invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0023] Figure 1 This is a schematic diagram of the irregular sand grain movement trajectory and attitude measurement scheme of the present invention, wherein (a) is a top view and (b) is a front view; Figure 2 This is a schematic diagram of the sand spreading device of the present invention, wherein (a) is the initial state and (b) is a schematic diagram of the sand particles moving with the airflow after the switching mechanism is opened; Figure 3 This is a schematic diagram of the sand particle identification and trajectory solving method of the present invention; Figure 4 This is a schematic diagram of obtaining the three-dimensional morphology of sand grains based on multi-view speckle images according to the present invention; Figure 5 This is a schematic diagram of the sand particle motion posture calculation method of the present invention, wherein (a) is the initial posture of the sand particle, and (b) is the instantaneous posture of the sand particle after motion and the posture change relative to the reference coordinate system.

[0024] Among them, 1-airflow generating device; 2-connecting pipe; 3-airflow direction; 4-sand grain spreading device; 5-fully developed zone; 6-impact observation zone; 7-drainage zone; 8-blade; 9-sand grain; 10-laser source; 11-shaping element; 12-laser illumination area; 13-camera; 14-switching mechanism; 15-upper wall; 16-lower wall. Detailed Implementation

[0025] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0026] It should be noted that the terms "first," "second," etc., in the specification and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0027] Example 1 This invention proposes a device for measuring the motion trajectory and attitude of irregular sand particles, including: a blade fixing device, an airflow pipe, an airflow generating device 1, a sand particle spreading device 4, a laser light source 10, a shaping element 11, a camera 13, and a data processing unit, etc.

[0028] During the experiment, sand particles 9 were carried along the airflow duct by the airflow (airflow speed of 50-200m / s) and collided with blades 8 at high speed in the impact observation area 6. The speckle image sequence of sand particles 9 was obtained by illumination by laser light source 10 and high-speed imaging by camera 13. Combined with motion trajectory extraction, two-dimensional morphology inversion and three-dimensional morphology reconstruction methods, the quantitative measurement of the motion trajectory and attitude change of irregular sand particles was realized.

[0029] The blade fixing device is used to stably fix the blade 8 to be tested (i.e., the turbine blade of the aero-engine) in a predetermined spatial position. It is not only used to ensure that the attitude of the blade 8 does not change during the test, but also serves as a spatial reference for measuring the trajectory and attitude of the sand particles 9.

[0030] The airflow duct is a hollow structure with a square cross-section, made of transparent material, which facilitates optical observation of the movement of sand particles inside the duct. The airflow duct is divided into three parts along the airflow direction: a fully developed zone 5, an impact observation zone 6, and a discharge zone 7.

[0031] The fully developed zone 5 is used to accelerate the sand particles 9 under the action of airflow and fully develop their motion state; the impact observation zone 6 is used to install the blade fixing device so that the sand particles 9 collide with the blade 8 in the impact observation zone 6; the discharge zone 7 is used to discharge the sand particles 9 and the debris generated after the impact from the airflow pipe to prevent residual particles from affecting subsequent tests and imaging quality.

[0032] The sand spreading device 4 is a cylindrical structure open at both ends. One end is connected to the starting end of the fully developed zone 5 of the airflow duct, and the other end is used to load sand particles 9. The sand spreading device 4 is equipped with a switching mechanism 14 at the end connected to the airflow duct. After the switching mechanism 14 is turned on before the test begins, the sand particles 9 in the sand spreading device 4 fall into the airflow duct under the action of gravity. The switching mechanism 14 can be set as a simple pull-out baffle.

[0033] The airflow generating device 1 is connected to the inlet of the airflow duct and is used to continuously supply airflow into the airflow duct, so that the sand particles 9 entering the airflow duct are carried by the airflow and accelerated. After the sand particles 9 are accelerated by the airflow and form a stable motion state in the fully developed zone 5, they enter the impact observation zone 6 and rush towards the surface of the blade 8 at a high speed, thus causing a high-speed impact.

[0034] The laser source 10 is used to generate a highly stable laser beam, and the shaping element 11 is used to shape the laser beam to form a laser illumination area 12 covering the impact observation area 6. Illuminated by the laser beam, the surface of the sand grains 9 entering the impact observation area 6 generates scattered light, thereby forming a clear speckle image at the imaging end.

[0035] Camera 13 is used for continuous high-speed imaging of the impact observation area 6. Camera 13 is fixedly mounted on the outside of the airflow duct, maintaining a constant viewing angle. It continuously captures speckle images of the sand particles 9 before and after they enter the impact observation area 6 and impact, forming a speckle image sequence. Each frame of the speckle image records corresponding time information (i.e., a timestamp), providing a time reference for subsequent motion trajectory and attitude analysis.

[0036] After the speckle image sequence was acquired, the speckle image sequence was first processed. The target detection and tracking algorithm was used to identify each sand grain 9 that entered the impact observation area 6, and the spatial position change information of the grain in the speckle image sequence was extracted, so as to obtain the motion trajectory of each sand grain 9 during the experiment.

[0037] The motion trajectory is used to characterize the spatial motion path of sand grain 9 before and after the impact, and serves as the basis for subsequent speckle image cropping and attitude measurement.

[0038] Based on the obtained motion trajectory of sand grain 9, the corresponding sand grain region is automatically cropped in each frame of the speckle image to obtain a speckle image sequence of each sand grain 9 at different times. Each speckle image sequence of sand grain 9 at different times is bound to the corresponding sand grain 9 and a timestamp to ensure data consistency and traceability.

[0039] Subsequently, the speckle image sequence is input frame by frame into a pre-trained generative network model. The generative network model is constructed based on a deep generative learning method and can invert the two-dimensional morphology image of sand grain 9 from the speckle image under the view of camera 13, thereby forming a two-dimensional morphology image sequence of each sand grain 9 at different times.

[0040] Since the sand grain 9 undergoes complex posture changes such as flipping and rotating during its movement, while the observation angle of the camera 13 remains fixed, the two-dimensional morphology images (i.e., the two-dimensional morphology image sequence) obtained by the same sand grain 9 at different times can be regarded as morphology information obtained from multiple equivalent perspectives.

[0041] Based on the above two-dimensional topography image sequence, a multi-view three-dimensional reconstruction method is used to perform three-dimensional topography inversion on sand grains 9, obtaining a three-dimensional topography model for each sand grain 9. Subsequently, based on the three-dimensional topography model of sand grain 9 and its corresponding two-dimensional topography image at each time step, the instantaneous attitude of sand grain 9 relative to a fixed viewpoint of camera 13 at each time step is determined.

[0042] By performing differential or derivative calculations on the instantaneous attitude of the same sand grain 9 within a time series, the rotation of the sand grain 9 in three spatial directions and the displacement of its center of mass within adjacent shooting intervals can be obtained, thereby enabling quantitative measurement of the motion trajectory and attitude change process of irregular sand grains.

[0043] Example 2 Reference Figure 1 The present invention proposes a device for measuring the motion trajectory and attitude of irregular sand particles, including an airflow generating device 1, a connecting pipe 2, a sand particle spreading device 4, an airflow pipe, a laser light source 10, and a camera 13.

[0044] The airflow generating device 1 is connected to the airflow pipe via the connecting pipe 2, and is used to introduce airflow into the airflow pipe. The airflow direction 3 is as follows: Figure 1 As shown in the diagram, the airflow duct is divided into a fully developed zone 5, an impact observation zone 6, and a discharge zone 7 along the airflow direction. The sand distribution device 4 is located at the beginning of the fully developed zone 5 and is used to introduce sand particles 9 into the airflow duct at the start of the test, so that the sand particles 9 move along the direction of the airflow duct under the action of the airflow.

[0045] Blade 8 (i.e., aircraft engine turbine blade) is installed in the impact observation area 6 of the airflow duct to collide with sand particles 9 moving with the airflow. The venting area 7 is used to discharge the impacted sand particles 9 and the debris generated by them from the airflow duct, preventing them from remaining in the impact observation area 6 and affecting subsequent tests.

[0046] The laser source 10 and the shaping element 11 are combined to generate and shape a laser beam, forming a laser illumination area 12 that covers the impact observation area 6, thereby illuminating the sand particles 9 entering the laser illumination area 12. A camera 13 is positioned outside the airflow duct and corresponds to the impact observation area 6. It is used to acquire the scattered light signal generated by the sand particles 9 within the laser illumination area 12, thereby obtaining a speckle image of the sand particles 9.

[0047] Reference Figure 2 A switching mechanism 14 is provided between the sand distribution device 4 and the airflow duct to control the timing of sand particles 9 entering the airflow duct. The airflow duct includes an upper wall 15 and a lower wall 16, both of which are made of transparent material to facilitate optical observation of the movement of sand particles 9 inside the airflow duct and their impact with the blades 8.

[0048] Figure 2 In diagram (a), the state before the test begins is indicated. At this time, the switch mechanism 14 is closed, and the sand particles 9 are confined within the sand particle spreading device 4 and have not yet entered the airflow duct. After the airflow generating device 1 is activated and a stable airflow is formed in the airflow duct, the switch mechanism 14 is opened. Under the action of gravity, the sand particles 9 fall from the sand particle spreading device 4 into the airflow duct and are carried along the direction of the airflow duct by the airflow, eventually entering the impact observation area 6 and colliding with the blade 8.

[0049] Figure 3 This is a schematic diagram of a method for identifying and solving the motion trajectory of sand particles. In this invention, the motion trajectory of sand particles is solved based on a sequence of speckle images acquired by camera 13 over a continuous time period.

[0050] In the speckle image captured at the first moment, the speckle image is first processed to identify each sand grain speckle in the image and calculate the centroid position corresponding to each sand grain speckle. A square search window is constructed with the centroid position as the center for subsequent matching.

[0051] Subsequently, in the speckle image captured at the second time step, the image region within the square search window constructed at the first time step is searched to identify the corresponding sand grain speckle and determine the centroid position of the sand grain speckle at the second time step. Based on the latest centroid position obtained at the second time step, the position and range of the corresponding square search window are updated.

[0052] Furthermore, in the speckle image captured at the third time point, the above search and update process is repeated to identify sand grain speckles within the updated square search window and obtain the centroid position of the sand grain speckles at the third time point. By repeating the above process in a continuous sequence of speckle images, continuous tracking of the same sand grain speckle at multiple time points is achieved.

[0053] By recording the centroid position of each sand grain at different times, and using the centroid position at the first time as a reference, connecting the centroid positions at each time in chronological order, the trajectory of the sand grain during the experiment can be obtained.

[0054] Figure 4 This is a schematic diagram of the process for obtaining the three-dimensional morphology of sand grains based on multi-view speckle images. In this invention, the acquisition of the three-dimensional morphology of sand grains is based on speckle images acquired by camera 13 at different times and the results of sand grain motion trajectory recognition.

[0055] First, based on the identification results of the sand grain movement trajectory, the speckle images acquired at different times are processed. The speckle images corresponding to the same sand grain 9 at each time are classified and collected to form a speckle image sequence for a single sand grain 9. Since the sand grain 9 will undergo posture changes such as flipping and rolling during its movement, the speckle images of the same sand grain 9 acquired at different times can be regarded as multi-view speckle images acquired from different equivalent viewpoints.

[0056] Subsequently, the multi-view speckle images obtained from the same sand grain 9 under different postures are used as input and imported into the pre-trained speckle to two-dimensional morphology image mapping model (i.e., generative network model) to process the speckle images and obtain the corresponding two-dimensional morphology images of the sand grain 9 under different views.

[0057] After obtaining the two-dimensional topographic image, edge detection processing is performed on the two-dimensional topographic image from various viewpoints to extract the two-dimensional contour information of sand grain 9 and generate a multi-view contour mask. Further, according to the imaging geometry of camera 13, the two-dimensional contour mask is back-projected into three-dimensional space to form multiple corresponding view frustums.

[0058] By performing spatial Boolean intersection operations on the aforementioned multiple view frustums, a three-dimensional envelope of sand grain 9 is obtained. Subsequently, the three-dimensional envelope is discretized to generate a corresponding three-dimensional voxel model. Through post-processing operations such as voxel smoothing and void filling, discontinuities on the surface and internal voids of the three-dimensional voxel model are eliminated, ultimately obtaining a three-dimensional morphology model of irregular sand grains.

[0059] Figure 5 This is a schematic diagram of the method for solving the sand grain attitude. In this invention, the sand grain attitude is solved based on the three-dimensional morphology model of the sand grain 9 obtained above and the two-dimensional morphology images corresponding to different times.

[0060] (a) Determination of reference attitude (first moment): First, the two-dimensional topography image obtained by inversion of the speckle image of sand grains captured at the first moment is used as the target image. A rotation transformation is applied to the three-dimensional topography model of the corresponding sand grains so that its forward two-dimensional projection under the current camera view gradually approaches and eventually matches the two-dimensional topography image.

[0061] The rotational transformation can be expressed as a transformation about the camera coordinate system. x , y , z The rotational combination of the three axes can be represented by the first rotation matrix R0. When the two-dimensional projection contour of the three-dimensional topography model and the target two-dimensional topography image meet the preset criteria in terms of contour overlap, principal axis direction and topography consistency, the rotational state is considered to be the attitude state of the sand grain at the first moment.

[0062] In this posture state, principal axis analysis is performed on the three-dimensional topography model to extract the major axis direction vector and minor axis direction vector of sand grain 9, and a reference coordinate system for sand grain 9 is constructed based on this, denoted as { O 0}( x , y , z The reference coordinate system is used as the reference coordinate system for subsequent attitude change calculations.

[0063] (ii) Solving for attitude at subsequent time points (second time point and beyond) The two-dimensional topography image obtained by inverting the speckle image of sand grains captured at the second moment is used as the new target image. Keeping the camera coordinate system unchanged, a new rotation and translation transformation is applied to the three-dimensional topography model of the same sand grain 9 so that its forward two-dimensional projection is consistent with the two-dimensional topography image at the second moment.

[0064] Once the consistency criterion is met, the major axis and minor axis direction vectors of the three-dimensional morphology model of the sand grains at that moment are extracted, and the corresponding current attitude coordinate system is constructed, denoted as { O 1}( x ', y ', z The current attitude coordinate system is relative to the reference coordinate system { O 0}( x , y , z The transformation relationship of ) can be expressed as rigid body pose transformation, including rotation transformation and translation transformation.

[0065] The rotation transformation can be represented by a rotation matrix:

[0066] In the formula, R0 and R1 represent the same sand grain 9 relative to the reference coordinate system at the first and second time points, respectively. O 0}( x , y , z The rotation matrices of R (i.e., the first rotation matrix and the second rotation matrix) 01 Let be the rotation matrix of sand grain 9 at the second moment relative to the first moment (i.e., the third rotation matrix).

[0067] Translation transformation can be represented by the difference between the centroid position vectors:

[0068] In the formula, t0 and t1 are the translation vectors of the same sand grain 9 relative to the origin (0,0,0) of the reference coordinate system at the first and second times, respectively (i.e., the first translation vector and the second translation vector), and Δt is the difference in the centroid position vector of sand grain 9 at the second time relative to the first time.

[0069] (III) Expression of attitude variables and attitude changes Based on the aforementioned third rotation matrix R 01 Furthermore, the attitude change of sand grain 9 can be expressed in any one of Euler angle form, axis-angle form, or quaternion form.

[0070] In a preferred embodiment, the third rotation matrix is ​​decomposed into rotations around... x , y , z The rotation of the three coordinate axes represents the change in attitude of sand grain 9 in the three directions: (Δ θ x ,Δ θ y ,Δ θ z ), used to describe the flipping, rolling, and deflection behavior of sand grain 9 between adjacent time points, where Δ θ x For sand particles 9 x The change in attitude in the direction, Δ θ y For sand particles 9 y The change in attitude in the direction, Δ θ z For sand particles 9 z The change in attitude in the direction.

[0071] By solving the above-mentioned attitude of the same sand grain 9 at each moment in a continuous time, and by performing differential calculation on the attitude change at adjacent moments, the attitude evolution process of sand grain 9 throughout the entire motion process can be obtained, including the change in rotation angle and the change in centroid displacement.

[0072] Example 3 Reference Figure 1 and Figure 2 This embodiment discloses a device for measuring the trajectory and attitude of irregular sand particles. The device mainly includes an airflow generating device 1, a sand particle spreading device 4, an airflow pipe, a blade fixing device, a laser illumination system (including a laser light source 10 and a shaping element 11), and a high-speed imaging system (camera 13).

[0073] (1) Installation and debugging of the test equipment like Figure 1 As shown, the airflow generating device 1 is connected to the inlet end of the airflow duct via the connecting pipe 2, and is used to provide a stable high-speed airflow into the airflow duct. The airflow direction 3 is as follows: Figure 1 As shown. The output flow rate and velocity of the airflow generating device 1 can be adjusted according to experimental requirements to simulate the sand flow environment under different operating conditions inside an aero-engine.

[0074] The airflow duct includes, in sequence along the airflow direction, a fully developed zone 5, an impact observation zone 6, and a discharge zone 7. The fully developed zone 5 is used to accelerate the sand particles 9 under the influence of the airflow and establish a stable motion state; the impact observation zone 6 is the core measurement area, used to arrange the tested blades and perform optical measurements; and the discharge zone 7 is used to promptly discharge the impacted sand particles and debris, preventing residual particles from interfering with subsequent experiments.

[0075] like Figure 1 and Figure 2 As shown, the sand distribution device 4 is positioned above the fully developed zone 5, with its lower end connected to the airflow duct for dispensing sand particles 9 into the airflow duct. A switching mechanism 14 is installed between the sand distribution device 4 and the airflow duct to control the timing of the sand particles 9 entering the airflow duct. Before the experiment begins, the switching mechanism 14 is in the closed state; after the airflow generating device 1 is activated and a stable flow field is formed, the switching mechanism 14 is opened, and the sand particles 9 fall under the influence of gravity and are carried into the airflow duct by the airflow.

[0076] The upper wall 15 and lower wall 16 of the airflow duct are made of transparent material to facilitate optical observation of the movement, attitude changes and impact process of the sand particles 9 inside the airflow duct.

[0077] The blade 8 is fixedly installed in the impact observation area 6. The spatial position and attitude of the blade 8 are locked by the blade fixing device so that it remains unchanged during the experiment, thus serving as a spatial reference benchmark for measuring the motion trajectory and attitude of the sand particles 9.

[0078] The laser source 10 is located outside the airflow duct. The laser beam emitted by the source is shaped by the shaping element 11 to form a laser illumination area 12 that covers the impact observation area 6. When the sand grains enter the laser illumination area 12, their rough surface structure will scatter the laser, thereby forming a speckle light field related to the morphology of the sand grains 9.

[0079] Camera 13 is positioned on one side of the laser illumination area 12, with its optical axis aligned with the impact observation area 6, for acquiring speckle images of sand grains 9 in a continuous time series. Camera 13 and laser light source 10 are synchronously triggered by a synchronization control system to avoid imaging blurring caused by the high-speed movement of sand grains 9.

[0080] After the above-mentioned device installation is completed, the airflow stability of the airflow generating device 1 is adjusted by the no-load ventilation method, and the imaging field of view of the camera 13 is spatially aligned with the laser illumination area 12 by static calibration, so that the impact observation area 6 is completely within the effective imaging range of the camera 13, thereby completing the device debugging.

[0081] (2) Method for measuring the trajectory and attitude of irregular sand particles based on the device of this embodiment After the installation and debugging of the above-mentioned device are completed, the irregular sand particle motion trajectory and attitude measurement is carried out based on the device of this embodiment. The specific operation process is as follows.

[0082] First, the airflow generating device 1 is activated, and airflow is introduced into the airflow pipe through the connecting pipe 2, so that the airflow flows sequentially through the fully developed zone 5, the impact observation zone 6 and the discharge zone 7 along the airflow flow direction 3, and forms a stable flow field inside the airflow pipe.

[0083] Subsequently, after the airflow stabilizes, the switch mechanism 14 is activated, and the sand particles 9 fall from the sand particle spreading device 4 into the airflow pipe. Under the combined action of gravity and high-speed airflow, they form a sand particle group and complete the acceleration process in the fully developed zone 5.

[0084] When the sand grains enter the impact observation area 6, the laser light source 10 forms a laser illumination area 12 under the action of the shaping element 11, illuminating the sand grains 9. Under synchronous triggering conditions, the camera 13 continuously acquires speckle images of the sand grains 9 at different times.

[0085] Based on the speckle images collected at different times, the sand grain speckles in the speckle images are first identified and matched (refer to...). Figure 3 By correlating the centroid positions of the sand grain speckle at continuous time intervals, the motion trajectory of each sand grain 9 before and after the impact is obtained.

[0086] Subsequently, based on the motion trajectory, the speckle images of the same sand grain 9 at different times were cropped and collected (refer to...). Figure 4The image is then input into a pre-trained speckle-to-two-dimensional topography image mapping model to obtain two-dimensional topography images of the sand grain 9 under different postures.

[0087] Furthermore, based on the two-dimensional morphology images of the same sand grain 9 under different postures, the three-dimensional morphology model of the sand grain 9 is reconstructed by multi-view contour back projection and spatial Boolean intersection operation (i.e., multi-view three-dimensional reconstruction method).

[0088] Finally, as Figure 5 As shown, the three-dimensional morphology model of sand grain 9 is projected and matched with the two-dimensional morphology images at different times to solve the instantaneous attitude of sand grain 9 at each time. By the change of instantaneous attitude between adjacent times, the rotation of sand grain 9 in three spatial directions (the three directions refer to the mutually perpendicular x, y, and z directions) and the displacement of the center of mass are obtained, thereby realizing the complete measurement of the motion trajectory and attitude of irregular sand grains.

[0089] Specifically, based on the acquisition and processing of speckle images of sand particles 9 by camera 13 in a continuous time series, the motion behavior and attitude changes of sand particles 9 in the impact observation area 6 are fully obtained through motion trajectory recognition, two-dimensional morphology reconstruction, three-dimensional morphology inversion and attitude solution.

[0090] a. Acquisition of speckle images and determination of motion trajectory for sand grain 9 1) Camera 13 is synchronously triggered within the laser illumination area 12 to continuously acquire speckle images of sand grains 9, forming a speckle image sequence. Each frame of speckle image is preprocessed, including background subtraction, grayscale normalization, and median filtering or wavelet denoising, to enhance the contrast between sand grain speckle and background and reduce high-frequency noise interference.

[0091] 2) In the speckle image at the first time step, all sand grain speckles are identified, and the centroid position of each sand grain speckle is calculated. A square search window is established with this centroid position as the center for speckle matching in subsequent time frames.

[0092] 3) In the speckle images at the second and subsequent time points, the square search window from the previous time point is used to search for sand grain speckles in the corresponding region and calculate the centroid position, while updating the position and range of the square search window. This process is repeated until all time frames are traversed, thereby obtaining the centroid position of each sand grain speckle at different time points.

[0093] 4) Connect the centroid positions of the same sand grain speckle in consecutive time frames in chronological order to form a complete motion trajectory, providing a spatial reference for subsequent two-dimensional and three-dimensional morphology reconstruction.

[0094] b. Three-dimensional morphology reconstruction of sand particles 1) Based on the motion trajectory of sand grain 9, the speckle images of the same sand grain 9 at different times are classified and collected to form a speckle image sequence of a single sand grain 9. Since the sand grain 9 will flip and roll during the motion, each frame of speckle image can be regarded as a multi-view speckle image obtained from different equivalent viewpoints.

[0095] 2) Input the above multi-view speckle images into a pre-trained speckle-to-2D topography image mapping model to generate corresponding 2D topography images of the sand grain 9 from different views. Obtain the 2D contour mask for each view through edge detection processing.

[0096] 3) Using the imaging geometry of camera 13, the two-dimensional contour mask is back-projected into three-dimensional space to generate the view frustum corresponding to each viewpoint. By performing a spatial Boolean intersection operation on all view frustums, the three-dimensional envelope of sand grain 9 is obtained.

[0097] 4) Discretize the three-dimensional envelope to generate a three-dimensional voxel model, and obtain a continuous and complete three-dimensional morphology model of irregular sand grains through post-processing operations such as smoothing filtering and void filling.

[0098] c. Solving for sand grain attitude and calculating attitude change 1) Determining the reference attitude (first moment) First, the two-dimensional topography image corresponding to the speckle image at the first moment is used as the target image. Second, a rotation transformation is applied to the three-dimensional topography model of sand grain 9 so that its forward projection in the camera coordinate system matches the two-dimensional topography image. Then, the rotation transformation is expressed as a revolution around... x , y , z The three-axis rotation combination is represented by the first rotation matrix R0. When the two-dimensional projected contour (i.e., the orthogonal projection) and the target two-dimensional topography image meet the preset criteria in terms of contour overlap, principal axis direction, and topography consistency, the rotation state corresponding to this rotation transformation is determined as the attitude state of sand grain 9 at the first moment, i.e., the reference attitude. Finally, the major axis vector and minor axis vector of sand grain 9 are extracted to construct the reference coordinate system { O 0}( x , y , z ).

[0099] 2) Attitude calculation at subsequent time steps (second time step and beyond) First, the two-dimensional topography image corresponding to the speckle image at the second time step is used as the new target image. A second rotation matrix R1 and a second translation vector t1 are applied to the three-dimensional topography model of the same sand grain 9 to make the two-dimensional projection consistent with the new target image. Second, the rotated major axis direction vector and minor axis direction vector are extracted to construct the current attitude coordinate system. O 1}( x ', y', z Then, the transformation relationship between the current attitude coordinate system and the reference coordinate system is expressed as a rigid body pose transformation, including a second rotation transformation (using the third rotation matrix R). 01 =R1·R0 - ¹ indicates translation transformation (using the difference in centroid position vector Δt=t1) t0 represents (the value of t0).

[0100] 3) Expression of attitude variables and temporal evolution First, the third rotation matrix R 01 It can be further represented in Euler angle form, axis-angle form, or quaternion form. In a preferred embodiment, it is decomposed into... x , y , z The rotation of the three coordinate axes (i.e., the attitude change in the three directions) is denoted as (Δ). θ x , Δ θ y , Δ θ z The method is used to describe the flipping, rolling, and deflection behavior of sand grain 9 between adjacent time points. Secondly, by solving the attitude of each frame at each time point in a continuous time period and calculating the attitude change between adjacent time points, the rotation of sand grain 9 in three spatial directions and the displacement of its center of mass are obtained, thus realizing the complete measurement of motion trajectory and attitude evolution.

[0101] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit its protection scope. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that after reading the present invention, they can still make various changes, modifications or equivalent substitutions to the specific implementation of the invention, but these changes, modifications or equivalent substitutions are all within the protection scope of the present invention pending approval.

Claims

1. A device for measuring the motion trajectory and attitude of irregular sand particles, characterized in that, It includes an airflow generating device (1), a connecting pipe (2), a sand spreading device (4), an airflow duct, a laser light source (10), a shaping element (11), a camera (13), and a data processing unit; The airflow generating device (1) is connected to the airflow pipe through the connecting pipe (2). The airflow pipe is divided into a fully developed zone (5), an impact observation zone (6), and a discharge zone (7) in sequence along the airflow direction (3). The sand grain spreading device (4) is set at the starting position of the fully developed zone (5) to introduce sand grains (9) into the airflow pipe and make the sand grains (9) move along the direction of the airflow pipe under the action of the airflow. The impact observation zone (6) is equipped with blades (8) to collide with the sand grains (9) moving with the airflow. The discharge zone (7) is used to discharge the impacted sand grains (9) and the debris generated by them from the airflow pipe. The laser light source (10) and the shaping element (11) are combined to generate a laser beam and shape the laser beam so that the laser beam forms a laser illumination area (12) covering the impact observation area (6), thereby illuminating the sand particles (9) entering the laser illumination area (12). The camera (13) is set outside the airflow duct and corresponds to the impact observation area (6) to collect the scattered light signal generated by the sand particles in the laser illumination area (12), thereby continuously capturing the speckle image of the sand particles (9) and forming a speckle image sequence. The data processing unit is used to obtain the motion trajectory of each sand grain (9) during the test process based on the speckle image sequence, obtain the two-dimensional morphology image sequence of each sand grain (9) at different times based on the motion trajectory of the sand grain (9), determine the instantaneous posture of the sand grain (9) at each moment relative to the fixed angle of the camera (13) based on the two-dimensional morphology image sequence, and obtain the rotation amount and centroid displacement of the sand grain (9) in three directions in space within the adjacent shooting interval based on the instantaneous posture, thereby realizing the quantitative measurement of the motion trajectory and posture change process of irregular sand grains; The process of obtaining the motion trajectory of each sand grain (9) during the experiment based on the speckle image sequence is as follows: The speckle image sequence is processed, and each sand grain (9) entering the impact observation area (6) is identified by the target detection and tracking algorithm. The spatial position change information of each sand grain (9) in the speckle image sequence is extracted, so as to obtain the motion trajectory of each sand grain (9) during the experiment. The process of obtaining a sequence of two-dimensional morphological images of each sand grain (9) at different times based on the movement trajectory of the sand grain (9) is specifically as follows: According to the motion trajectory of the sand grains (9), the corresponding sand grain region is automatically cropped in each frame of speckle image to obtain the speckle image sequence of each sand grain (9) at different times. The speckle image sequence at different times is input frame by frame into the pre-trained generative network model. The generative network model is built based on the deep generative learning method and can invert the two-dimensional morphology image of the sand grains (9) from the speckle image under the view of the camera (13), thereby forming the two-dimensional morphology image sequence of each sand grain (9) at different times. The determination of the instantaneous attitude of the sand grain (9) relative to a fixed viewing angle of the camera (13) at each moment based on the two-dimensional topographic image sequence is specifically as follows: Based on the two-dimensional topography image sequence, the three-dimensional topography of the sand grain (9) is inverted by the multi-view three-dimensional reconstruction method to obtain the three-dimensional topography model of each sand grain (9). Based on the three-dimensional topography model of the sand grain (9) and its corresponding two-dimensional topography image at each time, the instantaneous pose of the sand grain (9) relative to the fixed viewpoint of the camera (13) at each time is determined. The method of obtaining the rotation and centroid displacement of sand grains (9) in three spatial directions within adjacent shooting intervals based on instantaneous attitude is as follows: The instantaneous attitude of the same sand grain (9) in the time series is solved by difference or differentiation to obtain the rotation of the sand grain (9) in three directions in space and the displacement of the center of mass in adjacent shooting intervals.

2. The device for measuring the motion trajectory and attitude of irregular sand particles according to claim 1, characterized in that, The airflow duct is a hollow structure with a square cross-section, and the airflow duct is made of transparent material.

3. The device for measuring the motion trajectory and attitude of irregular sand particles according to claim 1, characterized in that, The impact observation area (6) is equipped with a blade fixing device, and the blade (8) is installed in the impact observation area (6) through the blade fixing device.

4. The device for measuring the motion trajectory and attitude of irregular sand particles according to claim 1, characterized in that, The sand spreading device (4) is a cylindrical structure with openings at both ends. One end of the cylindrical structure is connected to the starting end of the fully developed zone (5) of the airflow pipe, and the other end is used to load sand particles (9). The sand spreading device (4) is provided with a switching mechanism (14) for controlling the scattering of sand particles (9) at the end connected to the airflow pipe.

5. A method for measuring the motion trajectory and attitude of irregular sand particles, based on the device for measuring the motion trajectory and attitude of irregular sand particles as described in claim 1, characterized in that, include: Start the airflow generating device (1), and introduce airflow into the airflow pipe through the connecting pipe (2), so that the airflow flows through the fully developed zone (5), the impact observation zone (6) and the discharge zone (7) in sequence along the airflow direction (3), and forms a stable flow field inside the airflow pipe; The sand particles (9) are dropped into the airflow pipe by the sand particle spreading device (4), and a sand particle group is formed under the combined action of gravity and airflow, and the acceleration is completed in the fully developed zone (5); When the sand grains enter the impact observation area (6), the laser light source (10) forms a laser illumination area (12) under the action of the shaping element (11) to irradiate the sand grains (9). The camera (13) continuously acquires speckle images of the sand grains (9) at different times under synchronous triggering conditions. Based on the collected speckle images at different times, the sand grain speckles in the speckle images are identified and matched. By associating the centroid positions of the sand grain speckles at consecutive times, the motion trajectory of each sand grain (9) before and after the impact is obtained. According to the motion trajectory, the speckle images corresponding to the same sand grain (9) at different times are cropped and collected, and input into a pre-trained speckle to two-dimensional morphology image mapping model to obtain the two-dimensional morphology images of the sand grain (9) in different postures; based on the two-dimensional morphology images of the same sand grain in different postures, the three-dimensional morphology model of the sand grain (9) is reconstructed through a multi-view three-dimensional reconstruction method; specifically including: Based on the movement trajectory of the sand grain (9), the speckle images of the same sand grain (9) at different times are classified and collected to form a speckle image sequence of a single sand grain (9). Each frame of the speckle image sequence of a single sand grain (9) is a multi-view speckle image obtained from different equivalent viewpoints. The multi-view speckle image is input into the pre-trained speckle to two-dimensional topography image mapping model to generate the corresponding two-dimensional topography images of sand grains (9) under different views. The two-dimensional contour mask of each view is obtained by edge detection processing. Using the imaging geometry of the camera (13), the two-dimensional contour mask is back-projected into the three-dimensional space to generate the view frustum corresponding to each viewpoint. By performing spatial Boolean intersection operation on all view frustums, the three-dimensional envelope of the sand grain (9) is obtained. The three-dimensional envelope is discretized to generate a three-dimensional voxel model, and the three-dimensional morphology model of the sand grain (9) is obtained by smoothing filtering and void filling. The three-dimensional morphology model of sand grain (9) is projected and matched with the two-dimensional morphology images at different times to solve the instantaneous attitude of sand grain (9) at each time. By the change of instantaneous attitude between adjacent times, the rotation of sand grain (9) in three spatial directions and the displacement of the center of mass are obtained, thereby realizing the complete measurement of the motion trajectory and attitude of irregular sand grain; specifically: Using the two-dimensional topography image corresponding to the speckle image at the first moment as the target image, a rotation transformation and a first translation vector t0 are applied to the three-dimensional topography model of the sand grain (9). The rotation transformation is expressed as a revolution around the two-dimensional topography model of the sand grain (9). x , y , z The rotation combination of the three axes, the rotation transformation is represented by the first rotation matrix R0, so that its positive projection in the camera coordinate system is consistent with the two-dimensional shape image, then the rotation state corresponding to the rotation transformation is determined as the reference posture of the sand grain (9), the major axis vector and minor axis vector of the sand grain (9) are extracted, and the reference coordinate system is constructed. Using the two-dimensional topography image corresponding to the speckle image at the second moment as the new target image, the second rotation matrix R1 and the second translation vector t1 are applied to the three-dimensional topography model of the same sand grain (9) to make the two-dimensional projection consistent with the new target image. The major axis direction vector and minor axis direction vector after rotation are extracted to construct the current attitude coordinate system. The transformation relationship between the current attitude coordinate system and the reference coordinate system is expressed as rigid body pose transformation, including a second rotation transformation and a translation transformation. The second rotation transformation is performed using the third rotation matrix R1. 01 =R1·R0 - ¹ indicates that the translation transformation uses the centroid position vector difference Δt = t1. t0 represents; The third rotation matrix R 01 Decomposed into winding x , y , z The rotation of the three coordinate axes is denoted as (Δ). θ x , Δ θ y , Δ θ z ), used to describe the flipping, rolling and deflection behavior of sand grains (9) between adjacent moments. By solving the attitude of each frame at each moment in a continuous time and calculating the attitude change at adjacent moments, the rotation of sand grains (9) in three directions in space and the displacement of the center of mass are obtained, and the complete motion trajectory and attitude evolution measurement is realized.

6. The method for measuring the motion trajectory and attitude of irregular sand particles according to claim 5, characterized in that, Based on the collected speckle images at different times, the sand grain speckles in the speckle images are identified and matched. By associating the centroid positions of the sand grain speckles at consecutive times, the motion trajectory of each sand grain (9) before and after the impact is obtained, specifically: Preprocess the speckle images at different times; In the speckle image at the first moment after preprocessing, all sand grain speckles are identified and the centroid position of each sand grain speckle is calculated. A square search window is established with the centroid position as the center for speckle matching in subsequent time frames. In the speckle images at the second and subsequent time points, the square search window of the previous time point is used to search for sand grain speckles in the corresponding region and calculate the centroid position. At the same time, the position and range of the square search window are updated. This process is repeated until all time frames are traversed, so as to obtain the centroid position of each sand grain speckle at different time points. Connect the centroid positions of the same sand grain speckle in consecutive time frames in chronological order to form the motion trajectory of the sand grain (9) before and after the impact.

Images