Explosion fireball temperature field three-dimensional measurement method

Abstract

The invention discloses an explosion fireball temperature field three-dimensional measurement method, and relates to the field of explosive explosion temperature measurement. For a plane image obtained by multi-view shooting, an image gray gradient amplitude threshold segmentation method is adopted, a density-adjustable grid structure can be quickly constructed, and a three-dimensional temperature field of an explosion fireball is reconstructed under sparse projection data in combination with a joint algebraic reconstruction algorithm. The method is of great significance in acquiring explosion fireball temperature field information, deeply understanding a thermal damage mechanism acting on a target in the explosion process and constructing thermal damage protection measures.

Description

Technical Field

[0001] The present invention relates to the technical field of explosion temperature measurement, and in particular to a three-dimensional measurement method for the temperature field of an explosion fireball. Background Art

[0002] With the development of high-energy ammunition technologies such as thermobaric and fuel-air explosives, the thermal effects of temperature fields have played an important role in target damage, and the temperature of the explosion fireball has become one of the key parameters for measuring the damage capability of high-energy ammunition. Traditional temperature measurement techniques are mainly based on single-point measurement, but with the rapid development of science and technology, the limitations of these methods are becoming increasingly prominent when faced with the increasingly complex needs of future temperature measurement. The measurement of the transient explosion fireball temperature field is one of the core contents of studying the explosion mechanism and damage effectiveness of ammunition, and it is also an important link for in-depth analysis of the dynamic process of the explosion. To achieve the three-dimensional reconstruction of the transient explosion fireball temperature field, it is necessary to meet the rapid response requirements of transient high-temperature measurement and to collect projection data from multiple directions to provide the necessary support for the three-dimensional reconstruction of the temperature field.

[0003] Chinese patent 201910490137.2 discloses a method for reconstructing the three-dimensional temperature field of flames using a dual-light field camera. This method uses dual-view light field images to establish a cylindrical medium model and reconstructs the flame area grid based on effective light. However, the grid division structure of this method is fixed to a cylindrical shape, which lacks adaptability and makes it difficult to accurately depict complex temperature field structures such as transient explosion fireballs. In addition, its ability to respond to complex boundary changes is limited, which limits its reconstruction accuracy and versatility in irregular and highly dynamic thermal fields.

[0004] Chinese patent 202411182648.5 discloses a method for light field tomography and reconstruction of the three-dimensional temperature field of a flame. This method uses a light field imaging system to capture flame images and, combined with multi-point temperatures of thermocouples as a priori, constructs a set of radiation transfer equations to achieve the inversion of the three-dimensional temperature field of the flame. However, this method relies on precise ray tracing and radiation models, requires a large amount of radiation parameter calibration, and the system's pathological properties are still obvious in highly dynamic flame fields, making it difficult to meet the real-time, high-resolution measurement requirements of transient fireball temperature fields.

[0005] Chinese patent 202411944199.3 discloses a three-dimensional temperature field reconstruction method based on 3D Gaussian splattering. This method constructs an explicit Gaussian model to support three-dimensional reconstruction of temperature and color images through feature point extraction, sparse point cloud generation and differentiable rendering technology. However, this method is highly dependent on multimodal image training data and has weak adaptability to modal inconsistencies and noise disturbances in transient high-dynamic temperature fields. The reconstruction stability and controllability still need to be further improved. Summary of the Invention

[0006] In order to solve the above technical problems, a three-dimensional measurement method for the temperature field of an explosion fireball is provided, which solves the problem that the above-mentioned traditional tomography technology and reconstruction algorithm are difficult to meet the requirements of three-dimensional measurement of the temperature field of an explosion fireball.

[0007] In order to achieve the above objects, the technical solution adopted by the present invention is: A three-dimensional measurement method for the temperature field of an explosion fireball, characterized by comprising the following steps: Step 1: Using the image grayscale gradient amplitude threshold segmentation method, a grid structure with adjustable density can be quickly constructed; Step 2: Reconstruct the three-dimensional temperature field through the joint algebraic reconstruction algorithm.

[0008] The grid structure with adjustable density in step one provides a spatial carrier for three-dimensional measurement for the joint algebraic reconstruction algorithm in step two.

[0009] Compared with existing technologies, the advantages of the present invention are: Compared with traditional three-dimensional reconstruction methods that use fixed density or uniform voxel division, the present invention divides the response intensity according to the image grayscale gradient amplitude characteristics, automatically refines the grid in areas of drastic temperature changes, and merges or maintains sparse divisions in areas of gentle temperature changes, thereby significantly reducing computing resources and reconstruction time while ensuring local reconstruction accuracy. A joint algebraic reconstruction algorithm is used to reconstruct the temperature field of the explosion fireball in three dimensions. This algorithm has good adaptability to sparse projection data and is suitable for transient imaging scenarios with limited multi-viewpoints or incomplete data. Its iterative correction mechanism for projection errors can effectively suppress noise propagation and artifact generation, improving the numerical stability and edge resolution capabilities of the reconstruction, and has significant advantages in the reconstruction of the three-dimensional temperature field of the explosion fireball. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 This is a flow chart of step 1 in a method for three-dimensional measurement of the temperature field of an explosion fireball proposed by the present invention; Figure 2 This is a flow chart of step 2 in the method for three-dimensional measurement of explosion fireball temperature field proposed by the present invention; DETAILED DESCRIPTION

[0011] The present invention is described in further detail below with reference to the accompanying drawings.

[0012] Combine Figure 1 , the image grayscale gradient amplitude threshold segmentation method can be used to quickly construct a grid structure with adjustable density, which specifically includes the following steps: Step 1: Based on the distribution characteristics of the fireball radiation, determine the center of the three-dimensional intermediate region and construct a unified spatial coordinate system. For the explosion fireball temperature field, there is a camera coordinate system and an image coordinate system in each projection direction. Different cameras in the world coordinate system are represented using different camera coordinate systems. The camera coordinate origin is set to the intersection of the camera optical axis and the lens. By acquiring image data from different perspectives and combining it with the camera extrinsic parameter information, the corresponding rotation matrix and translation vector are calculated to obtain the mapping relationship between the world coordinate system, the camera coordinate system, and the image coordinate system.

[0013] Step 2-1: Calculate the grayscale gradient amplitude of the acquired image. The calculation formula is:

[0014]

[0015] Where, is the gradient image of the image in the direction, is the gradient image of the image in the direction, is the gradient magnitude.

[0016] Step 2-2: The calculated grayscale gradient amplitude can obtain a gradient amplitude image, and the Otsu method is used to find the optimal segmentation threshold. The Otsu method first performs histogram statistics on the gradient amplitude image and normalizes it into a probability distribution function to reflect the probability of occurrence of each grayscale gradient amplitude in the image. Subsequently, for all possible grayscale gradient amplitude thresholds, the image pixels are divided into two classes: one class is the background area with a grayscale gradient amplitude less than or equal to the grayscale gradient amplitude threshold, and the other class is the foreground area with a grayscale gradient amplitude greater than the grayscale gradient amplitude threshold. For each threshold, the pixel proportion and grayscale gradient amplitude mean of the two classes are calculated respectively, and the inter-class variance under the threshold is further calculated. The calculation formula for the inter-class variance is:

[0017] Where, is the proportion of values ​​less than or equal to the grayscale gradient amplitude threshold, is the proportion greater than the grayscale gradient amplitude threshold, is the mean value of the grayscale gradient amplitude that is less than or equal to the grayscale gradient amplitude threshold, It is the mean of the grayscale gradient amplitudes that are greater than the grayscale gradient amplitude threshold.

[0018] Find the threshold that maximizes the inter-class variance, which is the optimal partitioning threshold. Divide the reconstructed area into high-response areas and low-response areas based on the threshold.

[0019] Step 3: Refine the grid in the high response area, merge the grid or keep the sparse division in the low response area, build a grid structure with adjustable density, and divide the divided grid again.

[0020] Combine Figure 2 ,The 3D temperature field is reconstructed by the joint algebraic reconstruction algorithm, which specifically includes the following steps: Step 1: Capture the explosion fireball image at the same time from different viewing angles to obtain projection data at different angles. The projection data is the temperature value of each pixel in the image calculated from the grayscale value of the image pixel. The calculation formula is:

[0021] Where, 、 and It is the grayscale value of RGB three channels. 、 and is the corresponding wavelength, is the proportional coefficient.

[0022] Step 2: On the constructed adjustable density grid, a joint algebraic reconstruction algorithm is used to perform a three-dimensional inversion of the explosion fireball temperature field. First, an initial value is assigned to each voxel grid, which can be the average value of the entire field or a combination of prior information. Then, a joint algebraic reconstruction algorithm is used, and its iterative formula is:

[0023] Where, is the current reconstruction value, is the updated value, is the set of all rays under a certain projection angle, is the influence weight, is the actual projection value.

[0024] The iteration is based on whether the residual error of the projection data fitting converges and whether the overall change of the temperature field between two iterations is small enough to judge whether the optimal three-dimensional temperature distribution has been achieved.

[0025] Among them The calculation formula of the round residual is:

[0026] Where, is the projection matrix, is the current reconstruction value, is the projection data, and its convergence judgment condition is:

[0027] General Settings is 10-4 -10 -6 .

[0028] The calculation formula for the overall change of the temperature field in two iterations is:

[0029] The convergence judgment conditions are:

[0030] General Settings is 10 -3 -10 -5 .

[0031] Or when the iteration reaches the set maximum number of iterations, the iteration is forced to terminate and the final result is output.

Claims

1. A three-dimensional measurement method for the temperature field of an explosion fireball, characterized in that: The steps include: Step 1: Using the image grayscale gradient amplitude threshold segmentation method, a grid structure with adjustable density can be quickly constructed; Step 2: Reconstruct the three-dimensional temperature field through the joint algebraic reconstruction algorithm.

2. The method for three-dimensional measurement of the explosion fireball temperature field according to claim 1, characterized in that: The grid structure with adjustable density provides a spatial carrier for three-dimensional measurement for the joint algebraic reconstruction algorithm.

3. The method for three-dimensional measurement of the explosion fireball temperature field according to claim 1, characterized in that: Step 1 is as follows: Determine the center of the 3D reconstruction area, build a unified spatial coordinate system, and obtain the mapping relationship between the image and the reconstruction area; Calculate the grayscale gradient amplitude of the image, set the threshold and divide the reconstructed area into high response area and low response area; calculate the threshold in the high response area and low response area respectively, and then divide them into high response area and low response area respectively, until the temperature gradient is less than the test accuracy and the division stops; Refine the mesh in areas of high response and merge or keep the mesh sparse in areas of low response.

4. The method for three-dimensional measurement of explosion fireball temperature field according to claim 1, characterized in that: Step 2 is as follows: Collect multi-view explosion fireball images and obtain projection data at different angles; The optimal three-dimensional temperature distribution is obtained by iteratively solving the problem in a voxel grid based on a joint algebraic reconstruction algorithm.

Images