A method and system for reconstructing three-dimensional cell structure of detonation wave based on camera shooting

Abstract

The present application relates to the field of detonation diagnosis and optical measurement technology, and particularly relates to a detonation wave three-dimensional cell structure reconstruction method and system based on camera shooting. The method comprises the following steps: multiple high-speed cameras arranged circumferentially around a detonation tube are used to synchronously capture multi-angle two-dimensional images of a detonation wave front within a microsecond time window; detonation wave self-luminous imaging or laser-induced fluorescence enhancement imaging technology is used to collect more precise detonation wave structure; original images are subjected to contrast enhancement, noise suppression and optical distortion correction; a multi-view stereo vision algorithm and a motion compensation technology based on an optical flow method are combined to reconstruct a detonation wave transient three-dimensional density field, and three-wave point spatial topological trajectories are extracted from the three-dimensional density field. The present application overcomes the observation difficulty caused by image blurring due to high-speed motion; and ensures that the system can still stably acquire high spatio-temporal resolution three-dimensional data in a real harsh detonation environment.

Description

Technical Field

[0001] This invention relates to the field of detonation diagnosis and optical measurement technology, and in particular to a method and system for reconstructing the three-dimensional lattice structure of detonation waves based on camera-captured images. Background Technology

[0002] 3D reconstruction technology reconstructs the three-dimensional structure of a target object by acquiring two-dimensional image information from different perspectives. Existing 3D reconstruction technologies mainly include laser scanning reconstruction and camera-based reconstruction, and are widely used in fields such as explosion mechanics, combustion dynamics, and aerospace safety. However, laser scanning reconstruction equipment is expensive and complex to operate, making it unsuitable for reconstructing high-speed dynamic targets such as detonation waves. While traditional camera-based reconstruction methods are lower in cost and simpler to operate, they struggle to record the fine structure and instantaneous changes of cell lattices under the supersonic and unsteady propagation of detonation waves. This results in insufficient imaging timeliness and difficulty in extracting cell feature points, leading to low accuracy and poor structural integrity in the reconstructed 3D cell structure.

[0003] To address the need for reconstructing the three-dimensional cell structure of detonation waves, existing technologies employing single imaging methods have limitations in their applicability. When relying solely on self-luminescence imaging of detonation waves, uneven cell luminescence intensity or excessively high detonation wave propagation speeds fail to capture clear cell details and dynamic evolution characteristics. Relying solely on laser-induced fluorescence imaging requires complex laser emission and synchronization control equipment, increasing system cost and operational complexity, and making it difficult to adapt to scenarios with varying cell size changes under different initial pressures. Furthermore, existing multi-view imaging schemes lack scientific point placement and microsecond-level synchronous triggering mechanisms, making them prone to image temporal deviations due to the ultra-fast dynamic changes of detonation waves. This prevents accurate matching of cell features from multiple perspectives at the same moment, affecting the multi-view fusion effect and the final accuracy of cell structure reconstruction. Therefore, three-dimensional cell reconstruction technology urgently needs a system that is timely, targeted, highly accurate, and feasible to address the problems existing in current technologies. Summary of the Invention

[0004] To address the technical problem that existing technologies cannot acquire the transient three-dimensional spatial structure of detonation waves non-contactly and with high precision, this invention provides a method and system for reconstructing the three-dimensional lattice structure of detonation waves based on camera-captured images. The technical solution is as follows: On the one hand, a system for reconstructing the three-dimensional lattice structure of detonation waves based on camera imaging is provided. The system is characterized by comprising: a multi-view high-speed imaging unit, a detonation imaging enhancement unit, an enhanced full-transparency composite conduit unit, a multi-view synchronous triggering and control unit, an image acquisition and preprocessing unit, and a three-dimensional reconstruction and inversion unit. A multi-view high-speed imaging unit is used to acquire detonation wave data after the detonation experiment is triggered; it includes N high-speed cameras that are circumferentially and uniformly arranged at equal angular intervals on a rigid optical platform outside the detonation pipe. Each camera is equipped with a telecentric lens or a macro lens, and its optical axis is precisely intersected with the central axis of the pipe. The detonation imaging enhancement unit is used for active optical enhancement; it includes two modes: detonation wave autoluminescence imaging mode and laser-induced fluorescence enhancement imaging mode; it comprehensively judges the propagation speed of the detonation wave, cell luminescence intensity and reconstruction requirements, and executes one or a combination of the two modes. The enhanced fully transparent composite conduit unit serves as the main channel for detonation wave propagation and the core physical window for optical observation; the enhanced fully transparent composite conduit consists of a light-transmitting liner and a transparent reinforcing protective layer from the inside out; The multi-view synchronous triggering and control unit is used to receive the main trigger signal from the detonation experimental device, generate and distribute synchronous TTL pulse signals according to the preset delay parameters, and synchronously trigger the enhanced full-transparency composite pipe unit, the detonation imaging enhancement unit and the multi-view high-speed imaging unit; and physically connect the enhanced full-transparency composite pipe unit and the multi-view high-speed imaging unit. The image acquisition and preprocessing unit is used to connect to each camera in the multi-view high-speed imaging unit through a high-speed interface, receive and store raw image data to the solid-state disk array in real time, and configure image preprocessing algorithms. The 3D reconstruction and inversion unit is used to connect the image acquisition and preprocessing unit and is configured to perform 3D reconstruction and feature extraction algorithms on the preprocessed multi-view image sequences.

[0005] Optionally, the multi-view high-speed imaging unit also includes: The high-speed camera configuration includes: a global shutter, an adjustable single-frame exposure time range of 10 ns to 10 μs, a maximum frame rate of no less than 1 million frames per second, and, through synchronous trigger signal control, the simultaneous capture of multi-angle two-dimensional projection image sequences of the detonation wave front within the same microsecond-level time window when the detonation wave passes through the observation window.

[0006] Optionally, the detonation wave self-luminescence imaging mode includes utilizing the spontaneous emission generated by the chemical reaction zone of the detonation wave front, selectively acquiring chemiluminescence signals by installing narrowband filters with center wavelengths corresponding to specific free radical emission lines in front of each camera lens, and performing detonation wave imaging enhancement. The laser-induced fluorescence enhancement imaging mode includes a high-repetition-rate pulsed laser, a beam shaper, and a harmonic generator; it generates a sheet-like or planar laser beam that matches the absorption lines of the components in the flow field to be measured, thereby exciting specific molecules or atoms in the flow field to produce stimulated fluorescence, and the camera is equipped with a bandpass filter that matches the fluorescence wavelength for acquisition.

[0007] Optionally, the enhanced fully transparent composite duct unit also includes: The light-transmitting liner is made of optical-grade quartz glass or special transparent ceramic to form an observation area with high light transmittance; The transparent reinforced protective layer is made of quartz fiber or high-strength glass fiber wrapped around the outer wall of the light-transmitting liner in a specific winding manner and then bonded and cured with high-strength transparent resin.

[0008] Optionally, the multi-view synchronous triggering and control unit includes: During the system initialization phase, the delay time of each channel, including the laser emission delay, is preset by software. t laser Camera exposure delay t camera and exposure width t width ; Upon receiving the main experimental trigger signal, the atomic clock begins timing, and the delay generator operates at time [time value missing]. t laser Send a TTL trigger pulse with a pulse width of 10-20 ns to the laser; exist t camera At time 1, the delay generator simultaneously sends a pulse with a width of 1 to all N high-speed cameras. t width TTL exposure trigger pulse; For scenarios requiring higher temporal resolution, enable time-division exposure mode: the delay generator sends trigger pulses sequentially to N cameras, with the time interval between adjacent pulses being... Δt It can capture pseudo-video streams at an equivalent frame rate as low as 100 ns.

[0009] Optionally, the image acquisition and preprocessing unit is also used for: Background subtraction was performed on the static background image acquired before the experiment; For the image after background subtraction, noise suppression is performed based on nonlocal means or BM3D algorithm; After noise suppression processing, optical distortion correction and image alignment are performed based on camera calibration parameters. The corrected and aligned image is then subjected to adaptive image contrast enhancement based on a convolutional neural network.

[0010] Optionally, the 3D reconstruction and inversion unit is also used for: Based on epipolar geometry and an improved semi-global stereo matching algorithm, the disparity map between multi-view image pairs at each moment is calculated, and the three-dimensional point cloud data of the detonation wave front at that moment is reconstructed through the principle of triangulation. A motion estimation model based on the Horn-Schunck optical flow method is introduced to calculate the pixel-level velocity field in the image sequence and perform motion compensation on the 3D point cloud data to correct the data from all perspectives to the same absolute physical time. The motion-compensated multi-time 3D point cloud data is interpolated and fused into a unified 3D voxel grid to generate a continuous 3D density field distribution of detonation wave. In the three-dimensional density field, the density gradient modulus at each voxel location is calculated, the local density maxima region is identified as the Mach bar candidate region, the high density gradient ridge is extracted as the shock wave front, and the intersection point of the two in three-dimensional space is determined as the three-wave point. By tracing the spatial coordinates of three wave points along the time axis at consecutive moments, and using a Kalman filter or B-spline curve fitting algorithm, a smooth and continuous spatial motion trajectory of the three wave points is generated, thus obtaining the three-dimensional lattice structure of the detonation wave.

[0011] Optionally, the improved semi-global stereo matching algorithm includes: For the preprocessed and corrected left and right view images I L and I R Calculate its Census transform or gradient characteristics; Construct a disparity space image whose cost function C(p,d) combines the sum of absolute differences, gradient differences, and Hamming distance from the Census transform; The matching cost is aggregated along multiple paths (usually 8 or 16 directions) using dynamic programming, with the following formula:

[0012] , in, Indicates the cost of aggregation; Represents pixels p ; Indicates the disparity value; This represents the discrete disparity index within the disparity search range; P 1. P 2 is the penalty parameter. r For each path direction, the aggregated cost is summed, and a layer-by-layer comparison strategy is used to select the disparity value corresponding to the minimum aggregated cost for each pixel. d Generate an initial disparity map; The initial disparity map is post-processed by left-right consistency checking, sub-pixel interpolation, and weighted median filtering to obtain the final dense disparity map D( x,y ).

[0013] Optionally, three-wave point recognition includes: In a three-dimensional density field ρ ( x,y,z In ), the Sobel or Scharr operator is used to compute the value of each voxel in . x , y , z Density gradient in direction And calculate its modulus. ; When the high gradient threshold is T gradient ,Will > T gradient The voxels are labeled as candidate points of the shock wavefront, and continuous high-gradient ridges are extracted by the three-dimensional edge connection algorithm. Find local density maxima in the density field, satisfying... ρ ( x,y,z ) > ρ(neighbors) , representing the high-pressure zone of the Mach rod; In three-dimensional space, locate the points where high-gradient ridges intersect or are closely adjacent to local density maxima. Define these intersections or neighboring points as three-wave points and record their three-dimensional coordinates. X tp ,Y tp ,Z tp ).

[0014] On the other hand, a method for reconstructing the three-dimensional cell structure of detonation waves based on camera images is provided. This system is applied to a system for reconstructing the three-dimensional cell structure of detonation waves based on camera images. The method includes: S1. Arrange and precisely adjust the multi-view high-speed imaging unit, and use a high-precision calibration board to calibrate the internal and external parameters of each camera; acquire static background images and configure the delay parameters of the synchronous trigger and control unit; S2. After the detonation experiment is triggered, the synchronous triggering and control unit sequentially triggers the detonation imaging enhancement unit and all high-speed cameras to capture a set of multi-angle, high-temporal-resolution two-dimensional image sequences at the instant the detonation wave passes through the observation window. S3. Perform background subtraction, noise filtering, optical distortion correction, and contrast enhancement on the acquired raw image to obtain a high-quality calibrated image; S4. Using the calibration parameters, apply the stereo matching algorithm to the multi-view images at each moment to reconstruct the three-dimensional point cloud of the detonation wave front at that moment. S5. Calculate the optical flow field based on the image sequence, estimate the detonation wave velocity, and compensate for the deviation of the three-dimensional point cloud position caused by the camera exposure time difference or object motion, and align all data to a unified time reference. S6. Interpolate and fuse the motion-compensated 3D point cloud data from multiple consecutive moments into a unified 3D voxel grid to generate a 3D spatiotemporal density field for the detonation wave propagation process. S7. Automatically identify high gradient and high density regions from the three-dimensional density field, detect the spatial positions of the three wave points, and connect these points along the time axis to generate the spatial topological trajectory of the three-dimensional cell structure of the detonation wave.

[0015] On the other hand, a camera-based device for reconstructing the three-dimensional cell structure of detonation waves is provided. The camera-based device for reconstructing the three-dimensional cell structure of detonation waves includes: a processor; a memory, on which computer-readable instructions are stored. When the computer-readable instructions are executed by the processor, any one of the methods described above for reconstructing the three-dimensional cell structure of detonation waves based on camera images is implemented.

[0016] On the other hand, a computer-readable storage medium is provided, which stores at least one instruction, which is loaded and executed by a processor to implement any of the methods described above in the method for reconstructing the three-dimensional lattice structure of detonation waves based on camera images.

[0017] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following: 1. This invention achieves a leap from two-dimensional cross-sections to three-dimensional topological structural reconstruction, overcoming the dimensional limitations of traditional optical methods and contact recording methods. Traditional smoke trail methods can only leave two-dimensional projection information on the wall, failing to resolve the true three-dimensional spatial structure inside the detonation wave. This invention uses a circumferentially distributed multi-view high-speed imaging unit to capture a group of two-dimensional projections of the same transient state of the detonation wave under precise synchronization. The three-dimensional reconstruction and inversion unit, based on stereo vision principles and motion compensation algorithms, inverts the two-dimensional image sequence into a three-dimensional voxel density field, providing a direct experimental observation method for studying the spatial evolution, instability, and turbulent interactions of the lattice.

[0018] 2. This invention integrates dual-mode imaging of "passive self-luminescence" and "actively excited fluorescence." Traditional contact-type electrical probes can interfere with flow fields and are easily damaged, while single optical modes are often limited by specific conditions. The detonation imaging enhancement unit of this invention provides two adaptive modes: self-luminescence imaging mode and laser-induced fluorescence enhancement imaging mode. It can use the characteristic radiation of the detonation chemical reaction itself and the laser to actively excite specific components, either individually or simultaneously. This allows the system to perform both general structural observations and in-depth diagnosis of specific chemical processes, expanding the scientific detection dimensions of the system.

[0019] 3. Overcame the observation challenge of image blurring caused by high-speed motion. Under high-speed detonation conditions, achieving reliable optical observation faces the challenge of image blurring and reconstruction misalignment caused by the high-speed motion of the detonation wave. The 3D reconstruction and inversion unit introduces a motion compensation algorithm based on optical flow. By estimating and correcting pixel displacement and 3D point cloud misalignment caused by the high-speed motion of the wavefront and time-sharing exposure of multiple cameras, it ensures that the system can still stably acquire high spatiotemporal resolution 3D data under real and harsh detonation environments. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a block diagram of a three-dimensional lattice structure reconstruction system for detonation waves based on camera images provided by the present invention; Figure 2 This is a schematic diagram of the position arrangement of the multi-angle high-speed imaging unit provided by the present invention; Figure 3 This is a flowchart of a method for reconstructing a three-dimensional lattice structure of a detonation wave based on camera capture, provided by the present invention. Figure 4 This is a schematic diagram of the structure of the electronic device provided in an embodiment of the present invention. Detailed Implementation

[0022] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0023] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0024] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0025] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0026] This invention provides a system for reconstructing the three-dimensional cell structure of detonation waves based on camera-captured images. This system can be implemented using a camera-captured three-dimensional cell structure reconstruction device, which can be a terminal or a server. Figure 1 The diagram shows a block diagram of a three-dimensional cell structure reconstruction system based on camera-captured detonation waves. The system may include: a multi-view high-speed imaging unit 1, a detonation imaging enhancement unit 2, an enhanced full-transparency composite channel unit 3, a multi-view synchronous triggering and control unit 4, an image acquisition and preprocessing unit 5, and a three-dimensional reconstruction and inversion unit 6. The multi-view high-speed imaging unit 1 is used to acquire detonation wave data after the detonation experiment is triggered; it includes N high-speed cameras that are evenly arranged circumferentially at equal angular intervals on a rigid optical platform outside the detonation pipe. Each camera is equipped with a telecentric lens or a macro lens, and its optical axis is precisely intersected with the central axis of the pipe. Detonation imaging enhancement unit 2 is used for active optical enhancement; it includes two modes, namely detonation wave self-luminescence imaging mode and laser-induced fluorescence enhancement imaging mode; it comprehensively judges the propagation speed of the detonation wave, cell luminescence intensity and reconstruction requirements, and executes one or a combination of the two modes; The enhanced fully transparent composite pipe unit 3 is used as the main channel for detonation wave propagation and the core physical window for optical observation; the enhanced fully transparent composite pipe 3 includes a light-transmitting liner and a transparent reinforcing protective layer from the inside out; The multi-view synchronous triggering and control unit 4 is used to receive the main trigger signal from the detonation experimental device, generate and distribute synchronous TTL pulse signals according to the preset delay parameters; synchronously trigger the enhanced full-transparency composite pipe unit 3, the detonation imaging enhancement unit 2 and the multi-view high-speed imaging unit 1; and physically connect the enhanced full-transparency composite pipe unit 3 and the multi-view high-speed imaging unit 1. Image acquisition and preprocessing unit 5 is used to connect to each camera in multi-view high-speed imaging unit 1 through a high-speed interface, receive and store raw image data to solid-state disk array in real time, and configure image preprocessing algorithms. The 3D reconstruction and inversion unit 6 is used to connect the image acquisition and preprocessing unit 5 and is configured to perform 3D reconstruction and feature extraction algorithms on the preprocessed multi-view image sequence.

[0027] In one feasible implementation, the multi-view high-speed imaging unit 1 serves as the core data acquisition front-end of the system. N (N ≥ 4) high-speed cameras are circumferentially and uniformly arranged at equal angular intervals on a rigid optical platform outside the detonation pipe. Each camera is equipped with a telecentric lens or a macro lens, and their optical axes precisely intersect the central axis of the pipe. The high-speed cameras are configured with a global shutter, with a single-frame exposure time adjustable from 10 ns to 10 μs, a maximum frame rate of no less than 1 million frames per second, and are controlled by a synchronous trigger signal to simultaneously capture a sequence of multi-angle two-dimensional projection images of the detonation wave front within the same microsecond-level time window as the detonation wave passes through the observation window.

[0028] In one feasible implementation, the multi-view high-speed imaging unit 1 in this embodiment serves as the data acquisition front-end for the system to capture transient three-dimensional information. In laser-induced fluorescence (LIF) mode, its configuration must be strictly matched with the laser wavelength and fluorescence wavelength.

[0029] In terms of camera selection and configuration, at least four scientific-grade CMOS or enhanced high-frequency photography cameras with ultra-high frame rates and global shutter capabilities will be used. Each camera must have a single-frame resolution of no less than 1280 × 800 pixels, a frame rate of over 1 million frames per second at full resolution, and support external triggering and programmable exposure control. The minimum exposure time can be as short as 100 nanoseconds to record the detonation wave at the same moment. The camera sensor must have high quantum efficiency in the target fluorescence band (such as 308-310nm for OH radicals).

[0030] Regarding the lenses and optical accessories mounted on the cameras, each camera is equipped with an image-side telecentric lens whose depth of field must cover the entire diameter of the pipe to eliminate perspective errors and ensure consistent feature dimensions at different object distances. A narrow-band interference filter is mounted at the front of the lens, with its center wavelength precisely matched to the target fluorescence wavelength (e.g., OH). With a wavelength of 309 nm ± 2 nm and a bandwidth of 5-10 nm, it is used to suppress laser scattering light, continuous background radiation from detonation flames, and ambient stray light to the extreme, thereby improving the signal-to-noise ratio of the fluorescence signal by several orders of magnitude.

[0031] Regarding the spatial arrangement of the cameras, all cameras are uniformly arranged circumferentially around the observation section of the enhanced fully transparent composite pipe 3 at equal angular intervals using high-rigidity, finely adjustable supports. Figure 2 As shown. With the assistance of a laser tracker, the optical axes of all cameras are precisely aligned with the central axis of the pipe, forming a seamless stereoscopic vision network covering a 360° observation angle.

[0032] In one feasible implementation, the detonation imaging enhancement unit 2, as the active optical enhancement module of the system, is configured to operate in one or a combination of the following two modes, A and B, depending on the propagation speed of the detonation wave, the cell luminescence intensity, and the reconstruction requirements: Mode A: Detonation wave autoluminescence imaging mode. It utilizes the spontaneous emission generated by the chemical reaction zone of the detonation wave front. By installing narrowband filters with center wavelengths corresponding to specific free radical emission lines in front of each camera lens, it selectively collects the chemiluminescence signals of free radicals such as OH (309 nm) and CH (431 nm) to achieve the effect of detonation wave imaging enhancement. Mode B: Laser-induced fluorescence enhancement imaging mode, consisting of a high-repetition-rate pulsed laser, a beam shaper, and a harmonic generator, generates a sheet-like or planar laser beam that matches the absorption lines of the components in the flow field to be measured, thereby exciting specific molecules or atoms in the flow field to produce stimulated fluorescence. The high-speed camera in the multi-view high-speed imaging unit 1 is equipped with a bandpass filter that matches the fluorescence wavelength for acquisition.

[0033] In one feasible implementation, the specific configuration of the detonation imaging enhancement unit 2 in mode B is as follows: The laser is an optical parametric oscillator whose output wavelength can be tuned in the ultraviolet to visible light range to match the specific absorption line of the analyte. It can be used to measure OH radicals at wavelengths of 283 nm or 306 nm. When planar laser-induced fluorescence technology is used, the camera acquires the excited fluorescence signal; when particle image velocimetry technology is used, the system also includes tracer particles scattered in the flow field, laser sheet light irradiates the particles, and the camera records the Mie scattered light of the particles for subsequent velocity field calculation.

[0034] In one feasible implementation, the enhanced fully transparent composite pipe unit 3 serves as the physical basis for realizing circumferential unobstructed optical observation, as well as the main channel for detonation wave propagation and the core physical window for multi-view optical imaging. In LIF mode, the transmittance of ultraviolet laser and fluorescence needs to be additionally considered.

[0035] This invention abandons the traditional opaque metal pressure-bearing structure and adopts a fully transparent composite material solution, thereby forming an optical window that is transparent to visible light and selected electromagnetic radiation at all angles in the circumferential observation area. Structurally, the pipe consists of a light-transmitting liner that guides and seals the fluid and a transparent fiber-reinforced layer that provides explosion resistance and pressure resistance. The two are tightly bonded and cured with high-strength optically transparent resin, which together can withstand the transient impact pressure of tens of megapascals and the transient high temperature of thousands of Kelvin generated by the detonation wave, while ensuring that the transmission loss and wavefront distortion of the light signal when passing through the pipe wall are controlled within the range allowed by the system imaging.

[0036] As the inner structure that directly contacts the high-temperature and high-pressure reaction flow field, the light-transmitting liner focuses on the optical uniformity, surface shape accuracy, thermal shock resistance, and high transmittance characteristics within the target spectral range. The liner material is made of optical-grade fused silica glass or special transparent ceramics. These materials have extremely high transmittance (>90%) and thermal stability in the ultraviolet, visible, and near-infrared bands. They can efficiently transmit 283 nm excitation laser and 309 nm fluorescence, and can withstand the transient high temperature after the detonation wave front while maintaining optical performance. For specific experimental needs, the inner wall can be ultra-smoothly polished or coated with an anti-reflection film to maximize light transmission and reduce the interference of laser light reflection on the wall surface; the wall thickness of the liner is optimized through joint simulation of mechanical strength and optical distortion to minimize the thickness while meeting the pressure requirements, so as to reduce aberrations.

[0037] The transparent reinforced protective layer is made of quartz fiber or high-strength glass fiber wrapped around the outer wall of the light-transmitting liner in a specific winding manner, and then bonded and cured with high-strength transparent resin. Together, they enable the entire pipeline to maintain structural integrity and high optical transmittance when subjected to the high temperature and high pressure impact generated by the detonation wave.

[0038] The outer wall of the pipe is coated with a transparent fiber reinforcement layer using a precision winding process to provide the inner liner with the necessary mechanical strength and toughness to resist detonation impact. The material selected for this reinforcement layer must meet the requirements of high specific strength, high modulus, and a coefficient of thermal expansion that matches the inner liner. At the same time, the composite material itself must maintain excellent optical transparency after curing. Therefore, alkali-free glass fiber, quartz fiber, or special polymer optical fiber are selected as the reinforcing skeleton and impregnated in a high-strength transparent resin with a matching refractive index. Through an automated laying process that alternates between circumferential and helical directions, the fiber-resin composite layer is tightly wrapped and fixed to the outer wall of the transparent liner. This effectively constrains and counteracts the huge circumferential and axial stresses generated during the propagation of detonation waves. While ensuring that the physical structure of the pipe maintains complete sealing under extreme dynamic loads, it provides an ideal observation background for the externally arranged multi-view high-speed imaging unit 1, which is transparent across the entire field of view, free from visual obstruction, and with controlled optical distortion.

[0039] In this embodiment of the invention, a high-strength transparent fiber reinforcement layer is wrapped around the outside of the quartz liner using a wet winding process. The reinforcement fiber is selected from alkali-free S-glass fiber or pure quartz fiber with a coefficient of thermal expansion matching that of quartz, and is impregnated in highly transparent and highly tough epoxy resin or acrylic resin. A composite pressure-bearing structure is formed through alternating "circumferential + helical" winding. This layer maintains high transparency in the visible and near-ultraviolet wavelengths, ensuring that the line of sight of the external camera can penetrate the tube wall without distortion to observe the cross-section illuminated by the internal laser sheet.

[0040] In one feasible implementation, the multi-view synchronous triggering and control unit 4, as the system's precision timing hub, is crucial for achieving multi-view synchronous acquisition. It physically connects the enhanced fully transparent composite pipeline unit 3 and the multi-view high-speed imaging unit 1. Internally, it integrates a GPS atomic clock as the master clock source and is equipped with a multi-channel digital delay / pulse generator. Synchronization triggers 41 are installed on units requiring synchronous triggering, such as... Figure 1 As shown in (41).

[0041] This unit is configured to receive the main trigger signal from a detonation experiment (such as ignition) and, according to preset delay parameters... t 1 (Ensuring the laser is in place when the detonation wave reaches the laser sheet plane), generating and distributing synchronous TTL pulse signals with sub-nanosecond jitter accuracy, precisely triggering the exposure actions of the sensors in the enhanced full-transparency composite pipeline unit 3, the laser-induced fluorescence enhancement imaging module in the detonation imaging enhancement unit 2, and all cameras in the multi-view high-speed imaging unit 1 of the detonation imaging enhancement unit 2, ensuring that the propagation of the detonation wave and image acquisition are strictly synchronized in time and space.

[0042] After the laser pulse is emitted, and after a very short delay t2 (approximately equal to the time it takes for the laser pulse to travel to the observation area, typically tens of nanoseconds), the unit simultaneously sends exposure trigger pulses to all cameras in the multi-view high-speed imaging unit 1. The exposure time settings of all cameras must completely cover the laser pulse width and fluorescence lifetime.

[0043] To achieve higher temporal resolution, a camera time-sharing exposure mode can be enabled. In this mode, the unit sequentially sends trigger pulses to N cameras, with the interval between adjacent pulses as short as 100 nanoseconds, thereby obtaining an ultra-high frame rate pseudo-video sequence within a microsecond-level time window for more refined motion analysis.

[0044] In one feasible implementation method, the specific process is as follows: During the system initialization phase, the delay time of each channel, including the laser emission delay, is preset by software. t laser Camera exposure delay t camera and exposure width t width ; Upon receiving the experimental master trigger signal (rising edge), the atomic clock begins timing, and the delay generator operates at time [time value missing]. t laser A TTL trigger pulse with a pulse width of 10-20 ns is sent to the laser of the detonation imaging enhancement unit 2 to cause it to emit a laser pulse; exist t cameraAt time 1, the delay generator simultaneously sends a pulse with a width of 1 to all N high-speed cameras. t width TTL exposure trigger pulse; For scenarios requiring higher temporal resolution, enable time-division exposure mode: the delay generator sends trigger pulses sequentially to N cameras, with the time interval between adjacent pulses being... Δt It can be as low as 100 ns, thus achieving an equivalent frame rate of up to 1 / Δt Pseudo-video stream acquisition.

[0045] In one feasible implementation, the image acquisition and preprocessing unit 5 is connected to each camera in the multi-view high-speed imaging unit 1 via a high-speed Camera Link or CoaXPress interface, and is configured to receive and store raw image data to a solid-state drive array in real time. This unit is further configured to execute image preprocessing algorithms, including: background subtraction based on static background images acquired before the experiment, noise suppression based on non-local means or BM3D algorithms, optical distortion correction and image alignment based on camera calibration parameters, and adaptive image contrast enhancement based on convolutional neural networks.

[0046] The specific processing methods of the image acquisition and preprocessing unit 5 include: Background removal was performed on the static background image acquired before the experiment: The aim is to eliminate the constant background signal generated by the optical system itself, the fixtures, and the static environment, thereby highlighting the dynamic information caused solely by changes in the detonation wave flow field. Before the detonation experiment begins, the detonation imaging enhancement unit 2 is kept completely identical to the actual experiment, ensuring that the pipe is filled with a static, homogeneous unreacted gas mixture or air. All cameras are controlled to simultaneously acquire multiple frames (typically 100-500 frames) of images, and these images are time-domain averaged to greatly suppress random noise, resulting in a high signal-to-noise ratio static background image matrix. .in,( x,y () represents the pixel coordinates of the image.

[0047] During the detonation experiment, for each moment t Raw measurement images acquired Perform pixel-by-pixel subtraction in the complex domain (treating grayscale values ​​as complex numbers with an imaginary part of zero) or the real domain. This operation directly removes interference from camera dark current, fixed pattern noise, non-uniform background of laser sheet light, and static reflection from pipe walls, resulting in a difference image. This refers to the dynamic signal mainly generated by the detonation wave flow field modulation.

[0048] For the image after background removal, noise suppression is performed based on nonlocal means or the BM3D algorithm: Due to the output To smooth the image while preserving the true flow field features and edges, the image still contains random noise. I sub Applying nonlocal mean filtering to the difference image to be processed any pixel p its filtered value for: , From all pixels in the image q The weighted average determines the weights. w ( p,q ) depends on p and q The similarity between two image patches centered on a given point is not based on the degree of similarity in grayscale values ​​of the individual pixels. Here, Ω represents the similarity between two image patches centered on a given point. p The set of all pixels within the centered search window. This is a normalization constant. Weights w ( p,q )Depend on Defined in, Represented in pixels p The image patch vector centered on the image. This represents the square of the Gaussian-weighted Euclidean distance. a It is the standard deviation of the Gaussian kernel. h This is the attenuation parameter that controls the degree of filtering. This algorithm can effectively remove noise while preserving detailed features such as the shock front and reaction zone edges.

[0049] After noise suppression, optical distortion correction and image alignment are performed based on camera calibration parameters: This step forms the geometric foundation for subsequent multi-view 3D reconstruction. The goal is to correct and project the 2D images captured by all cameras onto the same coordinate system. Using methods such as Zhang Zhengyou's calibration, each camera is pre-calibrated with high precision to obtain its intrinsic and extrinsic parameters. Among these, the intrinsic parameter matrix... Including focal length f x , f y and principal point coordinates c x , c y This describes the camera imaging geometry. The distortion coefficient vector D typically contains radial distortion coefficients. k 1. k 2. k 3 and tangential distortion coefficient p 1. p 2. Used to describe lens nonlinear distortion. Extrinsic parameter matrix. For each camera, there is a 3 × 3 matrix R and a 3 × 1 translation vector. t This defines the pose of the camera coordinate system relative to a preset world coordinate system.

[0050] Using the intrinsic parameter matrix K and distortion coefficients D, the noise-suppressed image is processed. I denoised A nonlinear transformation is performed to correct the radial and tangential distortions introduced by the lens, resulting in an image without geometric distortion. I undistort .

[0051] Using the extrinsic matrix of each camera Calculate the first i Mapping each camera view to a homography matrix in a unified world coordinate system or directly reprojecting it. This involves mapping all camera views to a homography matrix in a unified world coordinate system. I undistort The image, based on its extrinsic parameters, is reprojected onto a virtual, common imaging plane located in the world coordinate system and perpendicular to the pipe axis. This process ensures that the same point in 3D space is reflected in all camera-corrected images. In this context, there is a strict and consistent geometric correspondence, which lays the foundation for stereo matching; The corrected and aligned image will undergo adaptive contrast enhancement based on a convolutional neural network: The image processed by the aforementioned steps I aligned The contrast between the flow field characteristics and the background may still be unsatisfactory under complex operating conditions. This step uses a deep learning-based method for adaptive enhancement.

[0052] A lightweight U-Net convolutional neural network is employed. This network aligns images with low contrast. I aligned As input, multi-scale features are extracted through an encoder-decoder structure, and skip connections are used to fuse shallow details and deep semantic information, ultimately outputting an image with significantly enhanced contrast. I enhanced .

[0053] The network is pre-trained on a large-scale synthetic detonation flow field image dataset. This dataset generates realistic flow field structures through computational fluid dynamics simulations and simulates different illuminations, smoke concentrations, and camera noise to generate corresponding "low-contrast-high-contrast" image pairs. Before deployment in a real system, the pre-trained model is fine-tuned using a small amount of real experimental data to improve its generalization ability under specific experimental conditions. During processing, I aligned By inputting each frame into the network, image sequences with clear features and high signal-to-noise ratio can be obtained. The final output is passed to the (6) 3D reconstruction and inversion unit 6.

[0054] In one feasible implementation, (6) the 3D reconstruction and inversion unit 6 is the core intelligent computing hub of this system. It receives the corrected and aligned multi-view image sequence output from the image acquisition and preprocessing unit 5 (5): , It executes a fully automated inversion algorithm chain from 2D images to 3D physical fields and topological features. This unit is typically implemented by a high-performance workstation or computing cluster equipped with multiple high-performance GPUs to handle computationally intensive tasks such as stereo matching and optical flow calculation. Its core algorithm flow is as follows: dense 3D reconstruction, motion compensation, spatiotemporal field fusion, and feature extraction.

[0055] The specific working steps of 3D reconstruction and inversion unit 6 are as follows: Semi-global stereo matching and dense 3D point cloud reconstruction: The goal of this step is to reconstruct the dense three-dimensional point cloud of the detonation wave surface from multiple viewpoint images at the same time t. P i ={( X,Y,Z,ρ )},in( X,Y,Z () is the world coordinate system. ρ The image intensity at that point. (Left view image) I L For reference, see the other N-1 right view images in sequence. I R A stereo image pair is formed, and an improved semi-global stereo matching algorithm is executed. For the reference image... I L any pixel p =( x,y and candidate parallax d Construct a matching cost function that combines multiple features. .in The sum of absolute differences measures color / grayscale similarity. It is a gradient difference and is robust to changes in illumination. The Hamming distance of the Census transform describes local structural similarity and is insensitive to noise. α , β These are weighting coefficients used to balance different cost terms.

[0056] To enhance the smoothness and consistency of the disparity map, along multiple (typically 8 or 16) path directions... r Perform dynamic programming to aggregate matching costs. Aggregate costs. Recursive update formula

[0057] ; in, Indicates the cost of aggregation; Represents pixels p ; Indicates the disparity value; This represents the discrete disparity index within the disparity search range; r This indicates the path direction. P 1 As a penalty for a parallax change of 1, P 2 To impose a greater penalty on parallax jumps (greater than 1), this design can smooth homogeneous regions while preserving object edges.

[0058] Sum of aggregation costs in all directions A layer-by-layer comparison strategy is used to select the pixel that makes... Minimum disparity value Generate an initial disparity map D 0( x , y Post-processing then follows: the left and right images are swapped and matched again, and inconsistencies (occlusions) are removed. (This is done in the context of parallax.) Quadratic curve fitting is performed on the vicinity to obtain sub-pixel precision disparity, improving depth resolution. Holes and noise points are filtered and filled to obtain the final, high-quality dense disparity map. D ( x , y According to camera calibration parameters (intrinsic parameter matrix) K ,base B Using the principles of triangulation, parallax is converted into three-dimensional coordinates. , , By fusing the reconstruction results of all stereo image pairs, a complete image with attributes at that moment is obtained. ρ 3D point cloud P t .

[0059] Motion estimation and point cloud time synchronization based on optical flow: To address the potential time synchronization issues caused by the high-speed motion of the detonation wave and time-sharing exposures by multiple cameras, this step performs motion compensation to correct all data to the same physical moment. t The Horn-Schunck optical flow method is employed on a two-dimensional image sequence at adjacent time points. This method minimizes... Energy functionals are used to solve for the globally smooth velocity field. u , v ).in, I x , Iy , I z It is the partial derivative of image brightness with respect to space and time. λ It is a regularization parameter. The velocity vector of each pixel is obtained by iteratively solving the Euler-Lagrange equation. u , v ).

[0060] For the projection time difference caused by multi-camera time-sharing exposure, it is assumed that the detonation wave is mainly along the pipe axis ( z (Axial) propagation, statistically analyzing the dominant motion direction from the optical flow field, and calculating the axial propagation velocity component. v z Let the first... i The offset of the exposure center of each camera relative to the global reference time t0 is obtained. Through step S 61 The camera has been reconstructed in t i 3D point cloud at time P i The axial displacement that needs to be compensated for at each point is The compensation formula is This operation is equivalent to bringing the observations from all cameras back to the same reference time. t Finally, all of them Transforming to a unified world coordinate system yields the following results: t Real-time synchronized, motion-blur-corrected complete 3D scene point cloud P synced ( t 0 ).

[0061] Spatiotemporal fusion and 3D density field generation: Multiple consecutive time points { t 0, t 1,..., t m Synchronized point cloud P synced ( t m By fusing the data, a continuous three-dimensional density field describing the propagation process of the detonation wave is constructed. ρ ( x , y , z In a coordinate system, define a cuboid region encompassing the entire detonation wave propagation path and discretize it into a regular three-dimensional voxel mesh. Each voxel is denoted as . V ijk Its center coordinates are ( x i , y j ,z k The size is set according to the required spatial resolution.

[0062] For each point in the motion-compensated 3D point cloud P n ( x n , y n , z n ) and its carried density attribute value ρ n Find the voxel containing that point. V ijk and its seven surrounding adjacent voxel vertices. ρ n Contributions, based on P n The distances to each voxel vertex are distributed to these 8 vertices using a trilinear interpolation algorithm. All point cloud data at all time points are traversed, and each voxel vertex accumulates contributions from multiple points. For each voxel... V ijk The final density value of the voxel center is obtained by weighting and averaging all the contribution values ​​received by its eight vertices. ρ ijk By traversing all voxels, a continuous three-dimensional density scalar field is generated. ρ ( x,y,z It can be used for isosurface extraction and visualization.

[0063] Automatic identification and tracking of three-point spatial topology trajectories: From the three-dimensional density field ρ ( x,y,z In the process, the key features of the three-wave point of the detonation wave lattice structure are automatically identified, and its trajectory is reconstructed. In the three-dimensional density field, the Sobel or Scharr operator is used to calculate the velocity of each voxel. x , y , z Density gradient in direction And calculate its modulus: ; Set a high gradient threshold T gradient ,Will > T gradient Voxel markers are used as shock candidate points, and a 3D edge-connection and skeletonization algorithm is employed to extract continuous high-gradient ridges, which represent the leading surfaces of the incident shock and transverse waves. In the density field, a search is conducted to find points that satisfy... ρ ( x,y,z ) > ρ(neighbors)The local density maxima are identified, corresponding to the Mach rod regions with the highest pressure within the cell. In three-dimensional space, locations are found where high-gradient ridges (shock fronts) and local density maxima regions (Mach rods) geometrically intersect or are spatially separated by a set tolerance. These spatial intersection points are designated as three-wave points, and their precise three-dimensional coordinates are recorded. X tp ,Y tp ,Z tp ).

[0064] Along the propagation direction of the detonation wave (i.e., the time axis), the three wave points identified in consecutive time steps (or consecutive 3D slices) are correlated and matched. Using a nearest neighbor algorithm, within a reasonable velocity constraint, the three wave points at the current time step are correlated with the three wave points closest to the next time step. The correlated discrete three-wave point spatial coordinate sequence is then smoothed using a Kalman filter. The Kalman filter, through a prediction-update loop, effectively suppresses measurement noise and generates an optimally estimated smooth motion trajectory. Alternatively, B-spline curves are used to fit the discrete points, generating continuous, smooth 3D spatial curves. Each curve represents the complete 3D spatial trajectory of a three-wave point during the detonation wave propagation process. The set of all trajectories constitutes a complete topological network describing the dynamic evolution of the 3D lattice structure of the detonation wave.

[0065] In this embodiment of the invention, to address the technical challenges of contact interference and dimension loss in existing detonation wave diagnostic techniques such as physical smoke trail methods, and the aliasing of projection information and inability to resolve three-dimensional spatial topology in traditional single-view optical imaging methods, this invention provides a camera-based system for reconstructing the three-dimensional cellular structure of detonation waves. This system utilizes multiple circumferentially arranged ultra-high-speed cameras to capture transient two-dimensional images of the detonation wave front from different angles under precise synchronization. More importantly, this system combines multi-view stereo vision algorithms with motion compensation technology based on optical flow to invert the two-dimensional image sequence into a three-dimensional density field, and automatically extracts the spatial evolution trajectories of three wave points, achieving non-contact and quantitative reconstruction of the fine three-dimensional cellular structure inside the detonation wave.

[0066] Figure 3 This is a flowchart illustrating a method for reconstructing a three-dimensional cell structure of a detonation wave based on camera imagery, according to an exemplary embodiment. This method is applicable to a system for reconstructing a three-dimensional cell structure of a detonation wave based on camera imagery. (Refer to...) Figure 3 The method includes: S1. Arrange and precisely adjust the multi-view high-speed imaging unit, and use a high-precision calibration board to calibrate the internal and external parameters of each camera; acquire static background images and configure the delay parameters of the synchronous trigger and control unit; S2. After the detonation experiment is triggered, the synchronous triggering and control unit sequentially triggers the detonation imaging enhancement unit and all high-speed cameras to capture a set of multi-angle, high-temporal-resolution two-dimensional image sequences at the instant the detonation wave passes through the observation window. S3. Perform background subtraction, noise filtering, optical distortion correction, and contrast enhancement on the acquired raw image to obtain a high-quality calibrated image; S4. Using the calibration parameters, apply the stereo matching algorithm to the multi-view images at each moment to reconstruct the three-dimensional point cloud of the detonation wave front at that moment. S5. Calculate the optical flow field based on the image sequence, estimate the detonation wave velocity, and compensate for the deviation of the three-dimensional point cloud position caused by the camera exposure time difference or object motion, and align all data to a unified time reference. S6. Interpolate and fuse the motion-compensated 3D point cloud data from multiple consecutive moments into a unified 3D voxel grid to generate a 3D spatiotemporal density field for the detonation wave propagation process. S7. Automatically identify high gradient and high density regions from the three-dimensional density field, detect the spatial positions of the three wave points, and connect these points along the time axis to generate the spatial topological trajectory of the three-dimensional cell structure of the detonation wave.

[0067] Figure 4 This is a schematic diagram of a device for reconstructing the three-dimensional lattice structure of detonation waves based on camera images, provided in an embodiment of the present invention. Figure 4 As shown, a device for reconstructing the three-dimensional lattice structure of detonation waves based on camera images may include the above-mentioned... Figure 1 The figure shows a system for reconstructing the three-dimensional lattice structure of detonation waves based on camera images.

[0068] Optionally, a camera-based device for reconstructing the three-dimensional lattice structure of detonation waves 410 may include a first processor 2001.

[0069] Optionally, a camera-based device for reconstructing the three-dimensional lattice structure of detonation waves 410 may further include a memory 2002 and a transceiver 2003.

[0070] The first processor 2001, memory 2002, and transceiver 2003 can be connected via a communication bus.

[0071] The following is combined Figure 4 A detailed description of each component of a camera-based device for reconstructing the three-dimensional lattice structure of detonation waves is provided below: The first processor 2001 is the control center of the camera-based detonation wave three-dimensional lattice structure reconstruction device 410. It can be a single processor or a collective term for multiple processing elements. For example, the first processor 2001 can be one or more central processing units (CPUs), application-specific integrated circuits (ASICs), or one or more integrated circuits configured to implement embodiments of the present invention, such as one or more digital signal processors (DSPs), or one or more field-programmable gate arrays (FPGAs).

[0072] Optionally, the first processor 2001 can perform various functions of a camera-based detonation wave three-dimensional lattice structure reconstruction device 410 by running or executing software programs stored in the memory 2002 and calling data stored in the memory 2002.

[0073] In a specific implementation, as one example, the first processor 2001 may include one or more CPUs, for example... Figure 4 CPU0 and CPU1 are shown in the diagram.

[0074] In a specific implementation, as one example, a camera-based device for reconstructing the three-dimensional lattice structure of detonation waves 410 may also include multiple processors, for example... Figure 4 The first processor 2001 and the second processor 2004 are shown in the diagram. Each of these processors can be a single-core processor or a multi-core processor. Here, a processor can refer to one or more devices, circuits, and / or processing cores used to process data (such as computer program instructions).

[0075] The memory 2002 is used to store the software program that executes the present invention, and is controlled by the first processor 2001 to execute it. The specific implementation method can be referred to the above method embodiment, and will not be repeated here.

[0076] Optionally, the memory 2002 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto. The memory 2002 may be integrated with the first processor 2001 or may exist independently, and may be connected via an interface circuit of a camera-based detonation wave three-dimensional lattice structure reconstruction device 410. Figure 4 (Not shown in the image) is coupled to the first processor 2001, and this embodiment of the invention does not specifically limit this.

[0077] The transceiver 2003 is used to communicate with network devices or with terminal devices.

[0078] Alternatively, transceiver 2003 may include a receiver and a transmitter. Figure 4 (Not shown separately). The receiver is used to implement the receiving function, and the transmitter is used to implement the transmitting function.

[0079] Optionally, the transceiver 2003 can be integrated with the first processor 2001 or exist independently, and can be connected to the interface circuit of a camera-based detonation wave three-dimensional lattice structure reconstruction device 410. Figure 4 (Not shown in the image) is coupled to the first processor 2001, and this embodiment of the invention does not specifically limit this.

[0080] It should be noted that, Figure 4 The structure of the device 410 for a method of reconstructing three-dimensional cell structure of detonation waves based on camera images shown does not constitute a limitation on the router. The actual knowledge structure recognition device may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0081] Furthermore, the technical effects of the camera-based method for reconstructing the three-dimensional cell structure of detonation waves can be referred to in the above-described method embodiments, and will not be repeated here.

[0082] It should be understood that the first processor 2001 in the embodiments of the present invention may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc.

[0083] It should also be understood that the memory in the embodiments of the present invention can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDR SDRAM), enhanced synchronous DRAM (ESDRAM), synchronous linked DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0084] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable sensors. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0085] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0086] It should be understood that, in various embodiments of the present invention, the order of the above-mentioned process numbers does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0087] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0088] 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.

[0089] 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.

[0090] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present 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, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention.

[0091] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A system for reconstructing the three-dimensional lattice structure of detonation waves based on camera-captured images, characterized in that, The system includes: a multi-view high-speed imaging unit, a detonation imaging enhancement unit, an enhanced full-transparency composite pipeline unit, a multi-view synchronous triggering and control unit, an image acquisition and preprocessing unit, and a three-dimensional reconstruction and inversion unit. A multi-view high-speed imaging unit is used to acquire detonation wave data after the detonation experiment is triggered; it includes N high-speed cameras that are circumferentially and uniformly arranged at equal angular intervals on a rigid optical platform outside the detonation pipe. Each camera is equipped with a telecentric lens or a macro lens, and its optical axis is precisely intersected with the central axis of the pipe. The detonation imaging enhancement unit is used for active optical enhancement; it includes two modes: detonation wave autoluminescence imaging mode and laser-induced fluorescence enhancement imaging mode; it comprehensively judges the propagation speed of the detonation wave, cell luminescence intensity and reconstruction requirements, and executes one or a combination of the two modes. The enhanced fully transparent composite conduit unit serves as the main channel for detonation wave propagation and the core physical window for optical observation; the enhanced fully transparent composite conduit consists of a light-transmitting liner and a transparent reinforcing protective layer from the inside out; The multi-view synchronous triggering and control unit is used to receive the main trigger signal from the detonation experimental device, generate and distribute synchronous TTL pulse signals according to preset delay parameters, synchronously trigger the enhanced full-transparency composite pipe unit, the detonation imaging enhancement unit and the multi-view high-speed imaging unit; and physically connect the enhanced full-transparency composite pipe unit and the multi-view high-speed imaging unit. The image acquisition and preprocessing unit is used to connect to each camera in the multi-view high-speed imaging unit through a high-speed interface, receive and store raw image data to the solid-state disk array in real time, and configure image preprocessing algorithms. The 3D reconstruction and inversion unit is used to connect the image acquisition and preprocessing unit and is configured to perform 3D reconstruction and feature extraction algorithms on the preprocessed multi-view image sequences.

2. The method according to claim 1, characterized in that, The multi-view high-speed imaging unit also includes: The high-speed camera configuration includes: a global shutter, an adjustable single-frame exposure time range of 10 ns to 10 μs, a maximum frame rate of no less than 1 million frames per second, and, through synchronous trigger signal control, the simultaneous capture of multi-angle two-dimensional projection image sequences of the detonation wave front within the same microsecond-level time window when the detonation wave passes through the observation window.

3. The method according to claim 1, characterized in that, The detonation wave self-luminescence imaging mode utilizes the spontaneous radiation generated by the chemical reaction zone of the detonation wave front. By installing narrowband filters with center wavelengths corresponding to specific free radical emission lines in front of each camera lens, the chemiluminescence signal is selectively acquired to enhance the detonation wave imaging. The laser-induced fluorescence enhancement imaging mode includes a high-repetition-rate pulsed laser, a beam shaper, and a harmonic generator; A sheet-like or planar laser beam is generated that matches the absorption lines of the components in the flow field to be measured, in order to excite specific molecules or atoms in the flow field to produce stimulated fluorescence. The camera is equipped with a bandpass filter that matches the fluorescence wavelength for acquisition.

4. The method according to claim 1, characterized in that, The enhanced fully transparent composite pipe unit also includes: The light-transmitting liner is made of optical-grade quartz glass or special transparent ceramic to form an observation area with high light transmittance; The transparent reinforced protective layer is made of quartz fiber or high-strength glass fiber wrapped around the outer wall of the light-transmitting liner in a specific winding manner and then bonded and cured with high-strength transparent resin.

5. The method according to claim 4, characterized in that, The multi-view synchronous triggering and control unit includes: During the system initialization phase, the delay time of each channel, including the laser emission delay, is preset by software. t laser Camera exposure delay t camera and exposure width t width ; Upon receiving the main experimental trigger signal, the atomic clock begins timing, and the delay generator operates at time [time value missing]. t laser Send a TTL trigger pulse with a pulse width of 10-20 ns to the laser; exist t camera At time 1, the delay generator simultaneously sends a pulse with a width of 1 to all N high-speed cameras. t width TTL exposure trigger pulse; For scenarios requiring higher temporal resolution, enable time-division exposure mode: the delay generator sends trigger pulses sequentially to N cameras, with the time interval between adjacent pulses being... Δt It can capture pseudo-video streams at an equivalent frame rate as low as 100 ns.

6. The method according to claim 1, characterized in that, The image acquisition and preprocessing unit is also used for: Background removal was performed on the static background image collected before the experiment; For the image after background removal, noise suppression is performed based on nonlocal means or BM3D algorithm; After noise suppression processing, optical distortion correction and image alignment are performed based on camera calibration parameters. The corrected and aligned image will be adaptively enhanced for contrast based on a convolutional neural network.

7. The method according to claim 6, characterized in that, The three-dimensional reconstruction and inversion unit is also used for: Based on epipolar geometry and an improved semi-global stereo matching algorithm, the disparity map between multi-view image pairs at each moment is calculated, and the three-dimensional point cloud data of the detonation wave front at that moment is reconstructed through the principle of triangulation. A motion estimation model based on the Horn-Schunck optical flow method is introduced to calculate the pixel-level velocity field in the image sequence and perform motion compensation on the 3D point cloud data to correct the data from all perspectives to the same absolute physical time. The motion-compensated multi-time 3D point cloud data is interpolated and fused into a unified 3D voxel grid to generate a continuous 3D density field distribution of detonation wave. In the three-dimensional density field, the density gradient modulus at each voxel location is calculated, the local density maxima region is identified as the Mach bar candidate region, the high density gradient ridge is extracted as the shock wave front, and the intersection point of the two in three-dimensional space is determined as the three-wave point. By tracing the spatial coordinates of three wave points along the time axis at consecutive moments, and using a Kalman filter or B-Spline curve fitting algorithm, a smooth and continuous spatial motion trajectory of the three wave points is generated, thus obtaining the three-dimensional lattice structure of the detonation wave.

8. The method according to claim 7, characterized in that, The improved semi-global stereo matching algorithm includes: For the preprocessed and corrected left and right view images I L and I R Calculate its Census transform or gradient characteristics; Construct a disparity space image whose cost function C(p,d) combines the sum of absolute differences, gradient differences, and Hamming distance from the Census transform; By aggregating matching costs along multiple paths using dynamic programming, the formula is: ， in, Indicates the cost of aggregation; Represents pixels p ; Indicates the disparity value; This represents the discrete disparity index within the disparity search range; P 1. P 2 is the penalty parameter. r Indicates the path direction; The aggregated costs of all paths are summed, and a layer-by-layer comparison strategy is used to select the disparity value corresponding to the minimum aggregated cost for each pixel. d Generate an initial disparity map; The initial disparity map is post-processed by left-right consistency checking, sub-pixel interpolation, and weighted median filtering to obtain the final dense disparity map D( x,y ).

9. The method according to claim 8, characterized in that, The three-wave point recognition includes: In a three-dimensional density field ρ ( x,y,z In ), the Sobel or Scharr operator is used to compute the value of each voxel in . x , y , z Density gradient in direction And calculate its modulus. ; When the high gradient threshold is T gradient ,Will > T gradient The voxels are labeled as candidate points of the shock wavefront, and continuous high-gradient ridges are extracted by the three-dimensional edge connection algorithm. Find local density maxima in the density field, satisfying... ρ ( x,y,z )> ρ(neighbors) , representing the high-pressure zone of the Mach rod; In three-dimensional space, locate the points where high-gradient ridges intersect or are closely adjacent to local density maxima. Define these intersections or neighboring points as three-wave points and record their three-dimensional coordinates. X tp ,Y tp ,Z tp ).

10. A method for reconstructing the three-dimensional cell structure of detonation waves based on camera-captured images, the method implementing the three-dimensional cell structure reconstruction system for detonation waves based on camera-captured images as described in any one of claims 1-9, characterized in that... The method steps include: S1. Arrange and precisely adjust the multi-view high-speed imaging unit, and use a high-precision calibration board to calibrate the internal and external parameters of each camera; acquire static background images and configure the delay parameters of the synchronous trigger and control unit; S2. After the detonation experiment is triggered, the synchronous triggering and control unit sequentially triggers the detonation imaging enhancement unit and all high-speed cameras to capture a set of multi-angle, high-temporal-resolution two-dimensional image sequences at the instant the detonation wave passes through the observation window. S3. Perform background subtraction, noise filtering, optical distortion correction, and contrast enhancement on the acquired raw image to obtain a high-quality calibrated image; S4. Using the calibration parameters, apply the stereo matching algorithm to the multi-view images at each moment to reconstruct the three-dimensional point cloud of the detonation wave front at that moment. S5. Calculate the optical flow field based on the image sequence, estimate the detonation wave velocity, and compensate for the deviation of the three-dimensional point cloud position caused by the camera exposure time difference or object motion, and align all data to a unified time reference. S6. Interpolate and fuse the motion-compensated 3D point cloud data from multiple consecutive moments into a unified 3D voxel grid to generate a 3D spatiotemporal density field for the detonation wave propagation process. S7. Automatically identify high gradient and high density regions from the three-dimensional density field, detect the spatial positions of the three wave points, and connect these points along the time axis to generate the spatial topological trajectory of the three-dimensional cell structure of the detonation wave.

Images