Virtual flame distance measurement method and system based on monocular vision and optical measurement

Abstract

The invention discloses a virtual flame distance measurement method and system based on monocular vision and optical measurement, which realize accurate measurement of flame distance and have the advantages of high measurement accuracy, low cost, simplicity and convenience in operation and the like. The patent model system comprises an image acquisition system, an image processing system, a distance calculation system and a display system. Wherein the image acquisition system is used for acquiring images of flames, the image processing system is used for processing the acquired images of the flames to obtain feature points and feature lines of the flames, the distance calculation system is used for calculating flame distances according to the position relation between the feature points and the feature lines, and the display system is used for displaying flame distance information. In the specific implementation, the device can extract and match the feature points and the feature lines of the collected flame image to obtain the three-dimensional coordinates of the flame, and further calculate the flame distance. Meanwhile, the device can adjust the measurement range and precision by changing parameters such as the focal length and the field of view, and the measurement accuracy is improved.

Description

Technical Field

[0001] The invention belongs to the field of flame distance measurement, and in particular relates to a virtual flame distance measurement method and system based on monocular vision and optical measurement. Background Art

[0002] In terms of real-time monitoring and evaluation of fire scenes, traditional fire monitoring methods mainly use sensors to monitor physical quantities such as temperature and smoke at the fire scene, and transmit the monitoring data to the monitoring center through data transmission devices for analysis and judgment. However, due to the complex environment of the fire scene, the characteristics of the flame color and shape will be affected by environmental factors, and sensor monitoring has limited limitations. At the same time, traditional fire monitoring methods cannot provide information on the distance of the flame and cannot accurately evaluate the flame burning situation.

[0003] In order to solve the above problems, some virtual flame distance measurement methods based on computer vision and optical measurement technology have been proposed in recent years. These methods use computer vision technology to identify and extract flame color, shape and other characteristics, use optical measurement technology to measure the flame distance, and use the distance information of virtual flames to monitor and evaluate the fire scene in real time.

[0004] Prior art document 1 (CN201523471U, device for controlling the color and brightness of electronic candles by ultrasonic ranging) discloses a device for controlling the color and brightness of electronic candles by ultrasonic ranging. The device uses the ultrasonic ranging principle to obtain the sound wave reflection formed by the palm above the electronic candle to measure the distance between the ultrasonic transducer and the palm reflection surface. However, the use of ultrasonic ranging has the problem of unstable measurement results and susceptibility to interference, which makes it difficult to meet the requirements of real-time monitoring and evaluation of the fire scene.

[0005] The monocular vision technology used in this application can effectively solve the accuracy and stability problems in optical measurement, and can measure flames at multiple angles. In the image processing, the collected flame image can be processed to obtain the characteristic points and characteristic lines of the flame. The distance calculation system provides more comprehensive flame distance information compared to the existing technology based on the positional relationship between the characteristic points and characteristic lines. For complex transformer fire scenes, more precise and anti-interference flame measurement technology is required. For this, a new virtual flame distance measurement method is needed, which can use monocular vision technology to improve measurement accuracy and stability, and provide more accurate flame distance information to better support real-time monitoring and evaluation of fire scenes.

[0006] Transformer fires are three-dimensional, hidden, and multi-scale, making it difficult to construct a dataset. The quality of the dataset directly affects the performance of the target detection model, so it is necessary to ensure that the dataset is both in line with the actual scene and can fully cover various situations of transformer fires. Monocular vision technology is more effective in constructing a multi-scene, realistic transformer fire flame dataset.

[0007] Transformer fires have unique flame morphology characteristics. According to the fire process, oil pool fires in transformer fires exist in all stages of transformer fires, while flowing fires and jet fires mainly exist in stable combustion and combustion mutation stages. Existing technologies are difficult to meet the needs of transformer fire image collection and measurement.

[0008] Prior art document 2 (CN 116935279A, application publication date 2023.10.24) discloses a method for identifying the development stage of a substation fire based on dynamic flame texture, including establishing a linear dynamic system model of the substation oil-filled equipment fire to characterize the dynamic characteristics of the texture and designing a kernel function based on the main angles of the subspace to measure the similarity of the flame dynamic model. Dynamic flame recognition is performed based on flame characteristics such as the flame spectrum and flicker. Although the method is simple, the algorithm has poor flexibility and often performs poorly for complex scenes. Although machine learning has made significant progress in flame detection, its ability to process complex data and large-scale data sets is relatively weak. This leads to certain limitations in the generalization performance and processing of highly nonlinear and complex relationships of machine learning methods in the field of flame detection.

[0009] The present invention regards the flame texture of oil-filled equipment fire that changes with time as a dynamic system, and couples the temporal dynamic features of the image sequence with the image features, and determines the development stage of the oil-filled equipment fire by establishing a generative model based on dynamic flame texture. It makes up for some defects of substation fire, that is, considering the nonlinear relationship between flame distance and brightness value, a new correction model is established using an exponential model, and the content of the nonlinear dynamic system model of substation oil-filled equipment fire is further improved. Summary of the invention

[0010] In order to solve the deficiencies in the prior art, the present invention provides a virtual flame distance measurement method and system based on monocular vision and optical measurement, especially a virtual flame distance measurement method suitable for real-time monitoring and evaluation of fire scenes. The present invention aims to provide a novel virtual flame distance measurement method that can improve measurement accuracy and stability and provide more accurate flame distance information, so as to better support real-time monitoring and evaluation of fire scenes.

[0011] The present invention adopts the following technical solution.

[0012] The present invention provides a virtual flame distance measurement method based on monocular vision and optical measurement, which is characterized by comprising:

[0013] Perform flame feature extraction, including flame edge detection, flame color space conversion and flame area segmentation;

[0014] Perform flame stereo reconstruction, including three-dimensional reconstruction of the flame area and three-dimensional reconstruction of the flame edge;

[0015] Perform flame distance correction, including the establishment of the correction model and the calculation of the distance correction coefficient.

[0016] Preferably, the flame feature extraction comprises:

[0017] Flame edge detection is to determine the shape and size of the flame by performing edge detection on the flame image, thereby realizing the extraction of flame features;

[0018] Extract the flame color and convert the image from RGB color space to HSV color space for processing;

[0019] Perform flame region segmentation, determine the threshold range of flame color, then extract all pixels in the image whose colors are within the threshold range, and finally optimize the extraction results through morphological processing;

[0020] The morphological processing includes dilation, erosion and closing operations.

[0021] Preferably, the flame stereoscopic reconstruction comprises:

[0022] Flame area extraction: flame feature extraction technology based on monocular vision is used, including flame edge detection, flame color space conversion and flame area segmentation, to obtain a two-dimensional image of the flame area;

[0023] Camera calibration: The camera is calibrated using the checkerboard calibration method to obtain the camera's intrinsic and extrinsic parameters;

[0024] 3D coordinate calculation: Using the camera's external parameters and the 2D coordinates of the camera and the flame area, the 3D coordinates of each pixel in the flame area are calculated, and the distance from the camera to the flame is calculated;

[0025] Point cloud reconstruction: Reconstruct the three-dimensional coordinate point cloud of the flame area to obtain a three-dimensional model of the flame area.

[0026] Preferably, the three-dimensional coordinate calculation includes:

[0027] The image coordinates are converted into three-dimensional coordinates in the world coordinate system using the camera's intrinsic and extrinsic parameters. The coordinates of the flame in the image are set to (u, v), and the corresponding world coordinates are set to (X, Y, Z).

[0028]

[0029] Among them, K represents the intrinsic parameter matrix of the camera, R and t represent the rotation matrix and translation vector of the camera respectively;

[0030] Set the Y and Z in the three-dimensional coordinates (X,Y,Z) to 0, that is, (X,0,0),

[0031]

[0032] Among them, r 11 ,r 12 ,r 13 Represent the elements of the rotation matrix R, t 1 ,t 2 ,t 3 Respectively represent the three components of the translation vector t;

[0033] Get the coordinates (u,v) of the three-dimensional coordinate (X,0,0) in the image and calculate the distance D from the flame to the camera:

[0034]

[0035] Among them, f is the focal length of the camera, and W is the width of the chessboard.

[0036] Preferably, the point cloud reconstruction process includes:

[0037] Data preprocessing: Import the 3D coordinate point cloud into the 3D reconstruction software and process the data according to the actual situation;

[0038] Point cloud registration: register point clouds from different perspectives and align point clouds taken at different positions into the same coordinate system;

[0039] Point cloud reconstruction: Use 3D reconstruction software to reconstruct the point cloud and generate a 3D model;

[0040] Model post-processing: post-process the generated 3D model to remove useless parts and cut the model;

[0041] Export model: Export the processed 3D model.

[0042] Preferably, the calibration model is

[0043] ln(D)=ln(a)+b·I′

[0044] Where D is the actual distance of the flame, a and b are correction coefficients, and I′ is the brightness value of the flame area.

[0045] Preferably, the calculation of the distance correction coefficient includes:

[0046]

[0047] Among them, i is the pixel, I′ is the brightness value of the pixel, n is the size of the data set, is the average value of I′, is the average value of ln(D).

[0048] The present invention also provides a virtual flame distance measurement system based on monocular vision and optical measurement, the system is a system used by the aforementioned virtual flame distance measurement method based on monocular vision and optical measurement, and is characterized in that it includes:

[0049] Flame feature extraction module, used for flame edge detection, flame color space conversion and flame area segmentation;

[0050] Flame stereo reconstruction module, used for 3D reconstruction of the flame area and the flame edge;

[0051] The flame distance correction module is used to establish a correction model and calculate the distance correction coefficient.

[0052] The present invention also provides a terminal, including a processor and a storage medium; the characteristics are:

[0053] The storage medium is used to store instructions;

[0054] The processor is configured to operate according to the instructions to execute the steps according to the aforementioned method.

[0055] The present invention also provides a computer-readable storage medium on which a computer program is stored, characterized in that the program implements the steps of the above method when executed by a processor.

[0056] The beneficial effect of the present invention is that, compared with the prior art, the present invention adopts a method combining computer vision technology and optical measurement technology, which can improve the precision and accuracy of measurement, and is suitable for measurement under different lighting conditions, both bright and dark environments. It provides a virtual flame distance measurement method based on computer vision and optical measurement technology, which has the advantages of high-precision measurement and small measurement error.

[0057] The present invention can be applied to various fields, especially the real-time monitoring and evaluation of fire scenes. The situation at the fire scene is complicated, and there are many fire-like obstacles that hinder identification or thick smoke in a dark environment that obscures the flame. In order to better learn the edge characteristics of the flame during the annotation process, the present invention does not arbitrarily segment the flame, but annotates the entire image contour including the flame while minimizing fire-like obstacles, and marks all multiple targets. The efficiency and accuracy of fire rescue work are improved. In addition, it can also be applied to animation, games and other fields, can provide a more realistic flame effect, and can more accurately control the size and position of the flame. Therefore, the present invention has broad application prospects and commercial value, and can be protected as a patent right. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 The figure is a flow chart of a virtual flame distance measurement method based on monocular vision and optical measurement in the present invention. DETAILED DESCRIPTION

[0059] In order to make the purpose, technical scheme and advantages of the present invention clearer, the technical scheme of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. The embodiments described in this application are only part of the embodiments of the present invention, not all of them. Based on the spirit of the present invention, other embodiments obtained by ordinary technicians in this field without creative work are all within the scope of protection of the present invention.

[0060] like Figure 1 As shown, embodiment 1 of the present invention provides a virtual flame distance measurement method based on monocular vision and optical measurement technology, especially a virtual flame distance measurement method suitable for real-time monitoring and evaluation of fire scenes.

[0061] Step 1: flame feature extraction technology based on monocular vision, including flame edge detection, flame color space conversion and flame area segmentation.

[0062] Step 1.1, flame edge detection is to determine the shape and size of the flame by performing edge detection on the flame image, thereby realizing the extraction of flame features. Common edge detection algorithms include Sobel operator, Canny operator and Laplacian operator. In this technical solution, an improved Sobel operator will be used for flame edge detection, which can more accurately extract the flame edge.

[0063] Step 1.2, flame color is one of the important indicators for extracting flame features, so it is necessary to convert the image from RGB color space to HSV color space for processing. HSV color space divides the image into three components: hue, saturation, and brightness, which can better reflect the changes in flame color. In this technical solution, the function in the OpenCV library is used to convert the image from RGB color space to HSV color space.

[0064] Step 1.3, flame region segmentation is the process of extracting the flame region from the image. This technical solution will adopt a segmentation algorithm based on color information. First, determine the threshold range of the flame color, then extract all the pixels in the image whose colors are within the threshold range, and finally optimize the extraction results through morphological processing. The morphological processing used in this technical solution includes expansion, erosion and closing operations. Through the above steps, we can realize the flame feature extraction technology based on monocular vision.

[0065] Step 2, flame stereo reconstruction technology based on monocular vision, including 3D reconstruction of the flame area and 3D reconstruction of the flame edge. The specific steps are as follows:

[0066] Step 2.1, flame area extraction: flame feature extraction technology based on monocular vision is used, including flame edge detection, flame color space conversion and flame area segmentation, to obtain a two-dimensional image of the flame area.

[0067] Preprocess the image, including smoothing, grayscale and other operations.

[0068] Using the flame edge detection and flame color space conversion technology in the core technology of step 1, the edge and color features of the flame are extracted.

[0069] The extracted results are subjected to morphological processing, including dilation, erosion and other operations, to obtain the flame area.

[0070] Step 2.2, camera calibration: Use the checkerboard calibration method to calibrate the camera to obtain the camera's intrinsic and extrinsic parameters.

[0071] Print a standard chessboard and fix it on a flat wall;

[0072] Use your camera to take 10-20 photos of the checkerboard at different positions and angles;

[0073] Use calibration software to calibrate these photos and obtain the intrinsic and extrinsic parameters of the camera.

[0074] Step 2.3, 3D coordinate calculation: Use the camera's external parameters and the 2D coordinates of the camera and the flame area to calculate the 3D coordinates of each pixel in the flame area and the distance from the camera to the flame. The specific steps are as follows:

[0075] Use a laser rangefinder or other distance measuring tool to measure the distance from the camera to the ground;

[0076] Measure the pixel size of the flame in the image;

[0077] The flame distance is measured using the camera's intrinsic and extrinsic parameters.

[0078] After completing the camera calibration and flame area extraction, the distance from the camera to the flame can be calculated based on the camera's intrinsic and extrinsic parameters and the position of the flame area in the image.

[0079] First, the image coordinates need to be converted into three-dimensional coordinates in the world coordinate system using the camera's intrinsic and extrinsic parameters. Assuming that the coordinates of the flame in the image are (u, v), its corresponding world coordinates are (X, Y, Z). It can be calculated using the following formula:

[0080]

[0081] Among them, K represents the intrinsic parameter matrix of the camera, R and t represent the rotation matrix and translation vector of the camera respectively.

[0082] Since we only need to calculate the distance from the flame to the camera, we can set Y and Z in the three-dimensional coordinates (X, Y, Z) to 0, that is, (X, 0, 0). At this time, the above formula can be simplified to:

[0083]

[0084] Among them, r 11 ,r 12 ,r 13 Represent the elements of the rotation matrix R, t 1 ,t 2 ,t 3 Represent the three components of the translation vector t respectively.

[0085] Solving the above equation, we can get the coordinates (u,v) of the three-dimensional coordinate (X,0,0) in the image. According to the principle of similar triangles, we can calculate the distance D from the flame to the camera:

[0086]

[0087] Where f is the focal length of the camera, and W is the width of the chessboard. Through the above method, the distance information from the camera to the flame can be obtained.

[0088] Step 2.4, point cloud reconstruction: reconstruct the three-dimensional coordinate point cloud of the flame area to obtain a three-dimensional model of the flame area.

[0089] Data preprocessing: Import the 3D coordinate point cloud into the 3D reconstruction software and process the data according to the actual situation, such as removing noise, smoothing the point cloud, etc.;

[0090] Point cloud registration: register point clouds from different perspectives, that is, align point clouds taken at different positions into the same coordinate system;

[0091] Point cloud reconstruction: Use 3D reconstruction software to reconstruct the point cloud and generate a 3D model. Common reconstruction methods include mesh reconstruction and voxel reconstruction. Mesh reconstruction is suitable for models with high requirements for details and surface performance, while voxel reconstruction is suitable for scenes with high requirements for model accuracy and model operability.

[0092] Model post-processing: Post-process the generated 3D model, such as removing useless parts, cutting the model, etc.

[0093] Export model: Export the processed 3D model to related software or equipment for use.

[0094] Step 3, flame distance correction technology based on monocular vision, including the establishment of a correction model and the calculation of a distance correction coefficient.

[0095] Step 3.1, Establishment of calibration model

[0096] In monocular vision, the measurement of flame distance is based on the relationship between brightness value and distance. In actual measurement, due to the influence of factors such as lighting and camera parameters under different conditions, measurement errors and inaccuracies will occur. Considering the nonlinear relationship between flame distance and brightness value, an exponential model is used to establish a correction model:

[0097] ln(D)=ln(a)+b·I′

[0098] Where D is the actual distance of the flame, a and b are correction coefficients, and I′ is the brightness value of the flame area.

[0099] Step 3.2, Calculation of distance correction coefficient

[0100] In order to obtain the parameters a and b in the correction model, we need to calculate through a set of known flame distances and corresponding image brightness values. The specific steps are as follows:

[0101] First, select a set of data sets, including known flame distances and corresponding image brightness values. These data can be obtained through on-site measurements or other reliable means. The size of the data set can be selected according to actual conditions, and it is usually recommended to contain at least 20 sets of data.

[0102] Using the flame feature extraction technology mentioned above, the pixel values ​​of the flame area are extracted from the image. Then, the average value of these pixel values ​​is calculated to obtain the average brightness value of the flame area.

[0103] Step 3.3, calculate the correction factor

[0104] Using the least squares method, we can find the estimated values ​​of a and b:

[0105]

[0106] Among them, i is the pixel, I′ is the brightness value of the pixel, n is the size of the data set, is the average value of I′, is the average value of ln(D).

[0107] Step 3.4, model calibration

[0108] The model is corrected using the estimated values ​​of a and b to obtain the final distance correction model:

[0109] ln(D)=ln(a)+b·I′+∈

[0110] Among them, ∈ is the noise error, which can be evaluated by methods such as cross-validation.

[0111] The distance correction coefficient is calculated as follows: The coordinates of the center point of the flame area measured by monocular vision are (u, v), then according to the camera model, the pixel coordinates of the point can be obtained Where (u c ,v c ) is the pixel coordinate of the camera optical center. According to the correction model, the estimated value of the flame distance D can be calculated by I′, that is, D = e ln(a)+b·I′ Furthermore, according to the ratio of the actual distance to the estimated distance, the distance correction coefficient can be obtained: in is the actual distance obtained by other sensors or manual measurement.

[0112] Through the above steps, the distance correction coefficient of monocular vision measurement can be obtained, thereby improving the measurement accuracy.

[0113] Embodiment 2 of the present invention provides a virtual flame distance measurement system based on monocular vision and optical measurement, the system is a system used by the aforementioned virtual flame distance measurement method based on monocular vision and optical measurement, and is characterized in that it includes:

[0114] Flame feature extraction module, used for flame edge detection, flame color space conversion and flame area segmentation;

[0115] Flame stereo reconstruction module, used for 3D reconstruction of the flame area and the flame edge;

[0116] The flame distance correction module is used to establish a correction model and calculate the distance correction coefficient.

[0117] The beneficial effect of the present invention is that, compared with the prior art, the present invention adopts a method combining computer vision technology and optical measurement technology, which can improve the precision and accuracy of measurement, and is suitable for measurement under different lighting conditions, both bright and dark environments. It provides a virtual flame distance measurement method based on computer vision and optical measurement technology, which has the advantages of high-precision measurement and small measurement error.

[0118] The present invention can be applied to various fields, especially the real-time monitoring and evaluation of fire scenes. The situation at the fire scene is complicated, and there are many fire-like obstacles that hinder identification or thick smoke in a dark environment that obscures the flame. In order to better learn the edge characteristics of the flame during the annotation process, the present invention does not arbitrarily segment the flame, but annotates the entire image contour including the flame while minimizing fire-like obstacles, and marks all multiple targets. The efficiency and accuracy of fire rescue work are improved. In addition, it can also be applied to animation, games and other fields, can provide a more realistic flame effect, and can more accurately control the size and position of the flame. Therefore, the present invention has broad application prospects and commercial value, and can be protected as a patent right.

[0119] The present disclosure may be a system, a method and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for causing a processor to implement various aspects of the present disclosure.

[0120] A computer-readable storage medium may be a tangible device that can hold and store instructions used by an instruction execution device. A computer-readable storage medium may be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. More specific examples of computer-readable storage media (a non-exhaustive list) include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanical encoding device, such as a punch card or a raised structure in a groove on which instructions are stored, and any suitable combination of the foregoing. As used herein, a computer-readable storage medium is not to be interpreted as a transient signal per se, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagating through a waveguide or other transmission medium (e.g., a light pulse through a fiber optic cable), or an electrical signal transmitted through a wire.

[0121] The computer-readable program instructions described herein can be downloaded from a computer-readable storage medium to each computing / processing device, or downloaded to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network can include copper transmission cables, optical fiber transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and / or edge servers. The network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in the computer-readable storage medium in each computing / processing device.

[0122] The computer program instructions for performing the operation of the present disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-related instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages, such as Smalltalk, C++, etc., and conventional procedural programming languages, such as "C" language or similar programming languages. Computer-readable program instructions may be executed completely on a user's computer, partially on a user's computer, as an independent software package, partially on a user's computer, partially on a remote computer, or completely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., using an Internet service provider to connect via the Internet). In some embodiments, an electronic circuit, such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA), may be customized by utilizing the state information of the computer-readable program instructions, and the electronic circuit may execute the computer-readable program instructions, thereby realizing various aspects of the present disclosure.

[0123] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A virtual flame distance measurement method based on monocular vision and optical measurement, characterized in that: include: Perform flame feature extraction, including flame edge detection, flame color space conversion and flame area segmentation; Perform flame stereo reconstruction, including three-dimensional reconstruction of the flame area and three-dimensional reconstruction of the flame edge; Perform flame distance correction, including the establishment of the correction model and the calculation of the distance correction coefficient.

2. The virtual flame distance measurement method based on monocular vision and optical measurement according to claim 1 is characterized in that: The flame feature extraction comprises: Flame edge detection is to determine the shape and size of the flame by performing edge detection on the flame image, thereby realizing the extraction of flame features; Extract the flame color and convert the image from RGB color space to HSV color space for processing; Perform flame region segmentation, determine the threshold range of flame color, then extract all pixels in the image whose colors are within the threshold range, and finally optimize the extraction results through morphological processing; The morphological processing includes dilation, erosion and closing operations.

3. The virtual flame distance measurement method based on monocular vision and optical measurement according to claim 2 is characterized in that: The flame stereo reconstruction comprises: Flame area extraction: flame feature extraction technology based on monocular vision is used, including flame edge detection, flame color space conversion and flame area segmentation, to obtain a two-dimensional image of the flame area; Camera calibration: The camera is calibrated using the checkerboard calibration method to obtain the camera's intrinsic and extrinsic parameters; 3D coordinate calculation: Using the camera's external parameters and the 2D coordinates of the camera and the flame area, the 3D coordinates of each pixel in the flame area are calculated, and the distance from the camera to the flame is calculated; Point cloud reconstruction: Reconstruct the three-dimensional coordinate point cloud of the flame area to obtain a three-dimensional model of the flame area.

4. The method for measuring virtual flame distance based on monocular vision and optical measurement according to claim 3 is characterized in that: The three-dimensional coordinate calculation includes: The image coordinates are converted into three-dimensional coordinates in the world coordinate system using the camera's intrinsic and extrinsic parameters. The coordinates of the flame in the image are set to (u, v), and the corresponding world coordinates are set to (X, Y, Z). Among them, K represents the intrinsic parameter matrix of the camera, R and t represent the rotation matrix and translation vector of the camera respectively; Set the Y and Z in the three-dimensional coordinates (X,Y,Z) to 0, that is, (X,0,0), Among them, r 11 ,r 12 ,r 13 They represent the elements of the rotation matrix R, t1, t2, t3 represent the three components of the translation vector t; Get the coordinates (u,v) of the three-dimensional coordinate (X,0,0) in the image and calculate the distance D from the flame to the camera: Where f is the focal length of the camera and W is the width of the chessboard.

5. The virtual flame distance measurement method based on monocular vision and optical measurement according to claim 4 is characterized in that: The point cloud reconstruction process includes: Data preprocessing: Import the 3D coordinate point cloud into the 3D reconstruction software and process the data according to the actual situation; Point cloud registration: register point clouds from different perspectives and align point clouds taken at different positions into the same coordinate system; Point cloud reconstruction: Use 3D reconstruction software to reconstruct the point cloud and generate a 3D model; Model post-processing: post-process the generated 3D model to remove useless parts and cut the model; Export model: Export the processed 3D model.

6. The virtual flame distance measurement method based on monocular vision and optical measurement according to claim 5 is characterized in that: The calibration model is ln(D)=ln(a)+b·I′ Where D is the actual distance of the flame, a and b are correction coefficients, and I′ is the brightness value of the flame area.

7. The virtual flame distance measurement method based on monocular vision and optical measurement according to claim 6 is characterized in that: The calculation of the distance correction coefficient includes: Among them, i is the pixel, I′ is the brightness value of the pixel, n is the size of the data set, is the average value of I', is the average value of ln(D).

8. A virtual flame distance measurement system based on monocular vision and optical measurement, the system being a system used in a virtual flame distance measurement method based on monocular vision and optical measurement as claimed in any one of claims 1 to 7, characterized in that: include: Flame feature extraction module, used for flame edge detection, flame color space conversion and flame area segmentation; Flame stereo reconstruction module, used for 3D reconstruction of the flame area and the flame edge; The flame distance correction module is used to establish a correction model and calculate the distance correction coefficient.

9. A terminal comprising a processor and a storage medium; characterized in that: The storage medium is used to store instructions; The processor is configured to operate according to the instructions to execute the steps of the method according to any one of claims 1-8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

Images