Fire Prevention Analysis System and Method Based on Multi-Source Data Fusion

Abstract

This invention relates to the field of fire prevention technology and discloses a fire prevention analysis system and method based on multi-source data fusion. The invention collects real-time video of the monitored area, identifies the dynamic evolution characteristics of smoke, analyzes smoke anomalies based on these characteristics, filters suspected thermal anomalies based on these anomalies, collects combustible data within a set range of the suspected thermal anomalies, analyzes the flammability of the combustibles based on this data, analyzes the combustion risk based on the smoke anomalies and combustible flammability of the suspected thermal anomalies, collects terrain and obstacle data within a set range of the suspected thermal anomalies, predicts oxygen entry under terrain conditions and obstacle obstruction based on this data, analyzes the probability of a fire occurring at the suspected thermal anomalies based on the combustion risk and oxygen entry, and determines whether a fire warning is needed based on the probability of a fire occurring at the suspected thermal anomalies, thereby improving the accuracy of fire warnings.

Description

Technical Field

[0001] This invention relates to the field of fire prevention technology, and in particular to a fire prevention analysis system and method based on multi-source data fusion. Background Technology

[0002] Fire threatens public safety and property security. The key to disaster prevention and mitigation lies in early fire prevention. Traditional fire detection technology mainly relies on point sensors, whose detection range is limited to the physical location of the sensor. There are many blind spots in large spaces, outdoor areas, or areas with complex structures. Moreover, sensors are easily affected by environmental pollution and airflow interference, resulting in delayed response and inability to assess the actual risk and development trend of a fire. Existing video smoke detection technology mainly analyzes the static features such as color and texture of a single frame image to identify the presence of smoke, but ignores the dynamic process of smoke generation and diffusion. This makes it difficult to distinguish fire smoke from visually similar interference such as water vapor, dust, and fog, resulting in false alarms. Furthermore, simple smoke detection ignores the physical properties of the environment in which it occurs and cannot combine environmental flammability to predict the spread of fire, resulting in limited value of early warning information.

[0003] Therefore, how to assess fire risks based on multi-source data fusion, accurately predict fire development trends, and provide rapid early warnings is a technical challenge that urgently needs to be solved in the field of fire prevention and control.

[0004] To address the aforementioned problems, this invention provides a fire prevention analysis system and method based on multi-source data fusion. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a fire prevention analysis system and method based on multi-source data fusion. This invention combines the dynamic evolution characteristics of smoke, the flammability of combustibles, and the oxygen supply situation. By comprehensively assessing the combustion risk and oxygen supply conditions, it determines the probability of fire occurrence, thereby improving the accuracy of fire early warning.

[0006] To achieve the above objectives, this invention provides a fire prevention analysis method based on multi-source data fusion, comprising the following specific steps:

[0007] S1. Collect real-time video of the monitored area, identify the dynamic evolution characteristics of smoke, analyze smoke anomalies based on the dynamic evolution characteristics of smoke, and screen suspected thermal anomaly points based on smoke anomalies.

[0008] S2. Collect combustible data within a set range of suspected thermal anomalies, and analyze the flammability of combustibles based on the combustible data;

[0009] S3. Analyze the combustion risk based on smoke anomalies and the flammability of combustibles at suspected thermal anomaly points;

[0010] S4. Collect terrain and obstacle data within a set range of suspected thermal anomalies, and predict the oxygen entry under terrain conditions and obstacle obstruction based on the terrain and obstacle data.

[0011] S5. Analyze the likelihood of a fire occurring at suspected thermal anomaly points based on the combustion risk and oxygen entry situation, and determine whether a fire warning is needed based on the likelihood of a fire occurring at suspected thermal anomaly points.

[0012] Preferably, step S1 includes the following specific steps:

[0013] S11. Deploy multiple high-definition cameras at key locations within the monitored area to simultaneously capture real-time video of the monitored area. Use the Gunnar Farneback algorithm to calculate dense optical flow and generate a two-dimensional motion vector for each pixel (x,y) in the image sequence. ,in, It is the velocity component in the x-direction. It is the velocity component in the y-direction;

[0014] S12. Extract the smoke regions from the image sequence, and obtain the average motion amplitude, motion consistency, region area growth rate, and texture change rate for each smoke region. The average motion amplitude is obtained using the average motion amplitude calculation formula, which is: In the formula, This represents the total number of pixels in the smoke area. Let be the motion vector of the i-th pixel in the smoke region. Motion consistency is obtained through the motion consistency calculation formula, which is: In the formula, Let be the direction angle of the motion vector of the i-th pixel in the smoke region. The average direction angle of all motion vectors within the smoke area is given. The area growth rate is obtained using the area growth rate calculation formula, which is: In the formula, Let be the smoke area at time t. A segmentation algorithm is used to perform pixel-level classification on the image at time t, generating a binary mask of the smoke. The smoke area is represented by 1, and the background by 0. The total number of pixels with a value of 1 in the mask is counted. The time interval is used to calculate the texture change rate using the texture change rate calculation formula, which is: In the formula, Let be the contrast at time t. The contrast is a texture feature based on the gray-level co-occurrence matrix, and the contrast calculation formula is: In the formula, u and v are the gray levels of the image. Let be the joint probability of gray levels u and v;

[0015] S13. Obtain the motion amplitude anomaly degree based on the ratio of the average motion amplitude to the fire smoke motion amplitude threshold, and obtain the smoke anomaly value by weighted summation based on the motion amplitude anomaly degree, motion consistency, regional area growth rate and texture change rate.

[0016] S14. Based on the comparison between the smoke anomaly value and the preset smoke anomaly threshold, if the smoke anomaly value is greater than or equal to the preset smoke anomaly threshold, then the smoke area corresponding to the smoke anomaly value is set as a suspected thermal anomaly point.

[0017] Preferably, step S2 includes the following specific steps:

[0018] S21. Collect combustible data within a set range of suspected thermal anomalies, wherein the combustible data includes the ignition point, volume and porosity of the combustible.

[0019] S22. Obtain combustible ignition point anomalies based on the difference between combustible ignition point and ambient temperature, obtain combustible volume anomalies based on the ratio of combustible volume to safe combustible volume threshold, and obtain porosity anomalies based on the ratio of porosity to safe porosity threshold.

[0020] S23. Obtain the flammability evaluation value of the combustible material by weighted summation of the combustible material ignition point anomaly value, combustible material volume anomaly value and porosity anomaly value.

[0021] Preferably, step S3 includes the following specific steps:

[0022] The combustion risk index is obtained by multiplying the smoke anomaly value of suspected thermal anomalies with the flammability evaluation value of combustibles.

[0023] Preferably, step S4 includes the following specific steps:

[0024] S41. Collect terrain data within a defined range of suspected thermal anomalies. The terrain data includes slope and curvature. The slope is obtained using a slope calculation formula, which is: In the formula, For elevation, Let X be the rate of change of elevation in the X direction. The elevation change rate in the Y direction is given by the curvature, which is obtained using the curvature calculation formula: In the formula, Let X be the second-order elevation change rate in the X direction. Let Y be the second-order elevation change rate in the Y direction;

[0025] S42. Obtain the slope influence value based on the ratio of the slope to the maximum slope in the region. By traversing the slope values ​​of all pixels, obtain the maximum slope in the scene as the maximum slope in the region. Obtain the curvature influence value based on the curvature normalization formula, which is: In the formula, This represents the minimum curvature of the region. To obtain the maximum curvature of the region, the curvature value of all pixels is traversed to obtain the maximum curvature value in the scene as the maximum curvature value of the region, and the minimum curvature value in the scene is obtained as the minimum curvature value of the region. The terrain influence factor is obtained by weighted summation of the slope influence value and the curvature influence value.

[0026] S43. Collect obstacle data within a set range of suspected thermal anomaly points. The obstacle data includes obstacle area, obstacle permeability, and obstacle spatial distribution coefficient. The obstacle area is the total projected area of ​​the obstacle. The obstacle permeability is obtained through the material porosity, where 1 indicates complete ventilation and 0 indicates complete obstruction. The obstacle spatial distribution coefficient is determined by path planning to determine the degree of overlap between the obstacle and the ventilation path, where 1 indicates that the obstacle is completely on the ventilation path and 0 indicates that the obstacle is not on the ventilation path.

[0027] S44. Calculate the obstacle influence factor based on the obstacle influence calculation formula, wherein the obstacle influence calculation formula is: In the formula, The area of ​​the obstacle. The obstacle spatial distribution coefficient. This refers to the area of ​​the obstacle along the ventilation path, that is, the area that actually blocks the airflow. For the ventilation rate of the obstacle, This represents the proportion of air actually blocked by the obstacle. The area of ​​the ventilation path. This represents the total proportion of obstacles that prevent oxygen from entering.

[0028] S45. The oxygen entry rate under terrain conditions and obstruction is obtained based on the product of the basic oxygen entry rate, terrain influence factor and obstacle influence factor. The basic oxygen entry rate is obtained by measuring the product of airflow and oxygen concentration in the air under ideal conditions.

[0029] Preferably, step S5 includes the following specific steps:

[0030] S51. Obtain the probability of a fire occurring at a suspected thermal anomaly point based on the product of the combustion risk index and the oxygen entry rate.

[0031] S52. Based on the comparison between the probability of a fire occurring at a suspected thermal anomaly point and a preset fire probability threshold, if the probability of a fire occurring at a suspected thermal anomaly point is greater than or equal to the preset fire probability threshold, it is determined that a fire warning needs to be issued.

[0032] This invention also provides a fire prevention analysis system based on multi-source data fusion, comprising:

[0033] The suspected thermal anomaly filtering module is used to collect real-time video of the monitored area, identify the dynamic evolution characteristics of smoke, analyze smoke anomalies based on the dynamic evolution characteristics of smoke, and filter suspected thermal anomalies based on smoke anomalies.

[0034] The combustible material flammability analysis module is used to collect combustible material data within a set range of suspected thermal anomaly points and analyze the flammability of combustible materials based on the combustible material data;

[0035] The combustion risk analysis module is used to analyze combustion risk based on smoke anomalies and the flammability of combustibles at suspected thermal anomaly points.

[0036] The oxygen entry analysis module is used to collect terrain data and obstacle data within a set range of suspected thermal anomalies, and to predict the oxygen entry situation under terrain conditions and obstacle obstruction based on the terrain data and obstacle data.

[0037] The fire early warning module is used to analyze the probability of a fire occurring at suspected thermal anomaly points based on combustion risk and oxygen entry, and to determine whether a fire early warning is needed based on the probability of a fire occurring at suspected thermal anomaly points.

[0038] The present invention also provides an electronic device, comprising: a processor and a memory, wherein the memory stores a computer program that can be called by the processor, and the processor executes the above-described fire prevention analysis method based on multi-source data fusion by calling the computer program stored in the memory.

[0039] The present invention also provides a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the above-described fire prevention analysis method based on multi-source data fusion.

[0040] The beneficial effects of this invention are as follows: This invention collects real-time video of the monitored area, identifies the dynamic evolution characteristics of smoke, analyzes smoke anomalies based on the dynamic evolution characteristics of smoke, screens suspected thermal anomalies based on smoke anomalies, collects combustible data within a set range of suspected thermal anomalies, analyzes the flammability of combustibles based on the combustible data, analyzes the combustion risk based on the smoke anomalies and combustible flammability of suspected thermal anomalies, collects terrain and obstacle data within a set range of suspected thermal anomalies, predicts oxygen entry under terrain conditions and obstacle obstruction based on terrain and obstacle data, analyzes the probability of fire at suspected thermal anomalies based on combustion risk and oxygen entry, and determines whether a fire warning is needed based on the probability of fire at suspected thermal anomalies. This invention combines the dynamic evolution characteristics of smoke, the flammability of combustibles, and the oxygen entry situation, and judges the probability of fire occurrence through a comprehensive assessment of combustion risk and oxygen supply conditions, thereby improving the accuracy of fire warning. Attached Figure Description

[0041] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the 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.

[0042] Figure 1 This is a schematic diagram of the fire prevention analysis method based on multi-source data fusion according to the present invention;

[0043] Figure 2 This is a schematic diagram of the S1 process of the fire prevention analysis method based on multi-source data fusion of the present invention;

[0044] Figure 3 This is a schematic diagram of the S4 process of the fire prevention analysis method based on multi-source data fusion of the present invention;

[0045] Figure 4 This is a schematic diagram of the fire prevention analysis system based on multi-source data fusion according to the present invention;

[0046] Figure 5 This is a schematic diagram of the electronic device structure of the present invention. Detailed Implementation

[0047] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention are within the scope of protection of the present invention.

[0048] Please see Figure 1 This invention provides a fire prevention analysis method based on multi-source data fusion, comprising the following specific steps:

[0049] S1. Collect real-time video of the monitored area, identify the dynamic evolution characteristics of smoke, analyze smoke anomalies based on the dynamic evolution characteristics of smoke, and screen suspected thermal anomaly points based on smoke anomalies.

[0050] Please see Figure 2 In this embodiment, S1 includes the following specific steps:

[0051] S11. Deploy multiple high-definition cameras at key locations within the monitored area to simultaneously capture real-time video of the monitored area. Use the Gunnar Farneback algorithm to calculate dense optical flow and generate a two-dimensional motion vector for each pixel (x,y) in the image sequence. ,in, It is the velocity component in the x-direction. It is the velocity component in the y direction. The Gunnar Farneback algorithm approximates the image neighborhood through polynomial expansion and uses the least squares method to calculate the pixel displacement.

[0052] S12. Preprocess the image sequence for denoising and enhancement, train the Faster R-CNN model to identify the rectangular bounding boxes of smoke, use the U-Net model to output pixel-level smoke masks, perform morphological operations on the segmentation results to obtain the smoke regions in the image sequence, extract the smoke regions from the image sequence, and obtain the average motion amplitude, motion consistency, region area growth rate, and texture change rate of each smoke region. The average motion amplitude reflects the intensity of the overall smoke motion and is obtained using the average motion amplitude calculation formula: In the formula, This represents the total number of pixels in the smoke area. Let be the motion vector of the i-th pixel in the smoke region. Fire smoke, driven by thermal buoyancy, moves much faster than other interfering objects, resulting in a larger average motion amplitude. Motion consistency reflects the degree of consistency in the directions of all vectors within the smoke region, and is obtained through the motion consistency calculation formula: In the formula, Let be the direction angle of the motion vector of the i-th pixel in the smoke region. The average direction angle of all motion vectors within the smoke area. Fire smoke is dominated by thermal buoyancy and has strong directional movement, thus exhibiting high uniformity in movement. The area growth rate reflects the speed of smoke diffusion and is obtained through the area growth rate calculation formula: In the formula, Let be the smoke area at time t. A segmentation algorithm is used to perform pixel-level classification on the image at time t, generating a binary mask of the smoke. The smoke area is represented by 1, and the background by 0. The total number of pixels with a value of 1 in the mask is counted. The time interval is used to measure the rapid spread of smoke during a fire, resulting in a high area growth rate. The texture change rate reflects the severity of changes in smoke texture and is obtained through the texture change rate calculation formula: In the formula, Let be the contrast at time t. Contrast is a texture feature based on the gray-level co-occurrence matrix, reflecting the spatial distribution difference between pixels with gray value u and pixels with gray value v. The greater the difference, the higher the contrast. The texture changes drastically in fire smoke, therefore the rate of change of contrast between adjacent frames is high. The contrast calculation formula is: In the formula, u and v are the gray levels of the image. If the image is an 8-bit grayscale image, the grayscale value range is 0-255, and u and v represent one of the grayscale values. The joint probability of gray levels u and v is obtained by the following steps: specify the distance and direction of the gray co-occurrence matrix, traverse all pixels in the image, count the number of pixel pairs that satisfy the condition that the current pixel has a gray value of u and the pixel has a gray value of v at the specified direction and distance, and obtain the joint probability based on the ratio of the number of pixel pairs to the total number of adjacent pairs. The total number of adjacent pairs can be the number of all possible adjacent pixel pairs in the image.

[0053] S13. Obtain the motion amplitude anomaly degree based on the ratio of the average motion amplitude to the fire smoke motion amplitude threshold, and obtain the smoke anomaly value by weighted summation based on the motion amplitude anomaly degree, motion consistency, regional area growth rate and texture change rate.

[0054] S14. Based on the comparison between the smoke anomaly value and the preset smoke anomaly threshold, if the smoke anomaly value is greater than or equal to the preset smoke anomaly threshold, then the smoke area corresponding to the smoke anomaly value is set as a suspected thermal anomaly point.

[0055] This embodiment uses multi-dimensional temporal features such as optical flow field analysis, texture evolution, and color change to accurately capture the diffusion, rising, and concentration change patterns of smoke. This not only effectively eliminates interference from non-fire smoke, but also reverse-screens out the source of risk based on the dynamic behavior patterns of smoke.

[0056] S2. Collect combustible data within a set range of suspected thermal anomalies, and analyze the flammability of combustibles based on the combustible data;

[0057] In this embodiment, S2 includes the following specific steps:

[0058] S21. Collect combustible material data within a set range of suspected thermal anomaly points. The combustible material data includes the ignition point, volume, and porosity of the combustible material. The steps for obtaining the volume of the combustible material are as follows: scan the combustible material with a lidar, generate point cloud data, and calculate the volume using 3D modeling software. The steps for obtaining the porosity are as follows: obtain the internal pore structure of the combustible material by scanning, calculate the pore volume and total volume by image segmentation, and obtain the porosity by the ratio of pore volume to total volume.

[0059] S22. Obtain combustible ignition point anomaly values ​​based on the difference between the combustible ignition point and the ambient temperature. The ambient temperature is obtained through a temperature sensor. The closer the ambient temperature is to the combustible ignition point, the higher the combustion risk. Obtain combustible volume anomaly values ​​based on the ratio of combustible volume to the safe combustible volume threshold. Obtain porosity anomaly values ​​based on the ratio of porosity to the safe porosity threshold.

[0060] S23. Obtain the flammability evaluation value of the combustible material by weighted summation of the combustible material ignition point anomaly value, combustible material volume anomaly value and porosity anomaly value.

[0061] S3. Analyze the combustion risk based on smoke anomalies and the flammability of combustibles at suspected thermal anomaly points;

[0062] In this embodiment, S3 includes the following specific steps:

[0063] The combustion risk index is obtained by multiplying the smoke anomaly value of suspected thermal anomalies with the flammability evaluation value of combustibles.

[0064] S4. Collect terrain and obstacle data within a set range of suspected thermal anomalies, and predict the oxygen entry under terrain conditions and obstacle obstruction based on the terrain and obstacle data.

[0065] Please see Figure 3 In this embodiment, S4 includes the following specific steps:

[0066] S41. Collect terrain data within a defined area of ​​suspected thermal anomalies. The terrain data includes slope and curvature. The slope is obtained using the slope calculation formula, which is: In the formula, Elevation is obtained through digital elevation models or 3D scanning data. Let X be the rate of change of elevation in the X direction. The elevation change rate in the Y direction can be calculated using the difference between adjacent pixels in a digital elevation model. Slope is the degree of inclination of the terrain, reflecting the direction and speed of airflow along the ground. Curvature is obtained through the curvature calculation formula, which is: In the formula, Let X be the second-order elevation change rate in the X direction. The second-order elevation change rate in the Y direction can be calculated using the second difference of a digital elevation model. Curvature is the degree of curvature of the terrain, reflecting the convergence or diffusion of airflow.

[0067] S42. Obtain the slope influence value based on the ratio of the slope to the maximum slope of the area. By traversing the slope values ​​of all pixels, obtain the maximum slope in the scene as the maximum slope of the area. The greater the slope, the faster the airflow speed along the ground, and the higher the oxygen entry rate. Obtain the curvature influence value based on the curvature normalization formula. The curvature normalization formula is as follows: In the formula, This represents the minimum curvature of the region. The maximum curvature of the region is obtained by iterating through the curvature values ​​of all pixels. The maximum curvature in the scene is taken as the maximum curvature of the region, and the minimum curvature in the scene is taken as the minimum curvature of the region. The positive or negative sign of curvature reflects the concavity and convexity of the terrain. Concave terrain (Q<0) gathers airflow and increases the oxygen entry rate, while convex terrain (Q>0) disperses airflow and reduces the oxygen entry rate. The terrain influence factor is obtained by weighted summation of the slope influence value and the curvature influence value. The terrain data indirectly changes the oxygen entry rate by affecting the path and speed of airflow.

[0068] S43. Collect obstacle data within the set range of suspected thermal anomaly points. The obstacle data includes obstacle area, obstacle permeability, and obstacle spatial distribution coefficient. The obstacle area is the total projected area of ​​the obstacle, which can be obtained through 3D vision. The obstacle permeability is the ability of the obstacle to allow air to pass through, which can be obtained through material porosity. 1 indicates complete ventilation and 0 indicates complete obstruction. The obstacle spatial distribution coefficient reflects the degree of obstruction of the obstacle on the ventilation path. 1 indicates that it is completely on the ventilation path and 0 indicates that it is not on the ventilation path. The degree of overlap between the obstacle and the ventilation path can be determined by path planning. The optimal ventilation path from the vent to the thermal anomaly point is found by path planning algorithm. The ratio of the projected area of ​​the obstacle on the path to the total area of ​​the obstacle is calculated.

[0069] S44. Calculate the obstacle impact factor based on the obstacle impact calculation formula. The obstacle impact calculation formula is as follows: In the formula, The area of ​​the obstacle. The obstacle spatial distribution coefficient. This refers to the area of ​​the obstacle along the ventilation path, that is, the area that actually blocks the airflow. For the ventilation rate of the obstacle, This represents the proportion of air actually blocked by the obstacle. The area of ​​the ventilation path. This represents the total proportion of oxygen entry blocked by obstacles. Obstacles reduce the oxygen entry rate by blocking airflow.

[0070] S45. The oxygen entry rate under terrain conditions and obstruction is obtained by multiplying the basic oxygen entry rate, terrain influence factor and obstacle influence factor. The basic oxygen entry rate is the oxygen volume flow rate under ideal conditions, which reflects the rate at which oxygen enters through the vent when there is no terrain or obstacle obstruction. It is obtained by measuring the product of air flow rate and oxygen concentration in the air under ideal conditions.

[0071] S5. Analyze the likelihood of a fire occurring at suspected thermal anomaly points based on the combustion risk and oxygen entry situation, and determine whether a fire warning is needed based on the likelihood of a fire occurring at suspected thermal anomaly points.

[0072] In this embodiment, S5 includes the following specific steps:

[0073] S51. Obtain the probability of a fire occurring at a suspected thermal anomaly point based on the product of the combustion risk index and the oxygen entry rate.

[0074] S52. Based on the comparison between the probability of a fire occurring at a suspected thermal anomaly point and a preset fire probability threshold, if the probability of a fire occurring at a suspected thermal anomaly point is greater than or equal to the preset fire probability threshold, it is determined that a fire warning needs to be issued.

[0075] The steps for obtaining the weights and thresholds in this embodiment are as follows: acquire real-time video of the historical monitoring area, analyze historical suspected thermal anomalies, historical flammability of combustibles, historical combustion risk, and historical oxygen entry rate, import the acquired historical data into each step of this embodiment to obtain the probability of fire occurring at suspected thermal anomalies and the judgment results on whether a fire warning is needed, and simultaneously acquire historical actual fire occurrences. Import the judgment results and historical actual fire occurrences together into MATLAB fitting software for fitting, and obtain the set of weights and thresholds with the highest judgment accuracy.

[0076] Please see Figure 4 The present invention also provides a fire prevention analysis system based on multi-source data fusion, including: a suspected thermal anomaly point screening module, used to collect real-time video of the monitoring area, identify the dynamic evolution characteristics of smoke, analyze smoke anomalies based on the dynamic evolution characteristics of smoke, and screen suspected thermal anomalies based on smoke anomalies.

[0077] The combustible material flammability analysis module is used to collect combustible material data within a set range of suspected thermal anomaly points and analyze the flammability of combustible materials based on the combustible material data;

[0078] The combustion risk analysis module is used to analyze combustion risk based on smoke anomalies and the flammability of combustibles at suspected thermal anomaly points.

[0079] The oxygen entry analysis module is used to collect terrain data and obstacle data within a set range of suspected thermal anomalies, and to predict the oxygen entry situation under terrain conditions and obstacle obstruction based on the terrain data and obstacle data.

[0080] The fire early warning module is used to analyze the probability of a fire occurring at suspected thermal anomaly points based on combustion risk and oxygen entry, and to determine whether a fire early warning is needed based on the probability of a fire occurring at suspected thermal anomaly points.

[0081] Please see Figure 5 The present invention also provides an electronic device, including a processor and a memory, wherein the memory stores a computer program that can be called by the processor, and the processor executes the above-described fire prevention analysis method based on multi-source data fusion by calling the computer program stored in the memory.

[0082] The electronic device can vary considerably depending on its configuration or performance. It may include one or more Central Processing Units (CPUs) and one or more memories, wherein the memory stores at least one computer program, which is loaded and executed by the processor to implement the fire prevention analysis method based on multi-source data fusion provided in the above-described embodiment. The electronic device may also include other components for implementing its functions; for example, it may have wired or wireless network interfaces and input / output interfaces for data input and output. Details will not be elaborated upon in this embodiment.

[0083] This invention also provides a computer-readable storage medium storing instructions that, when a computer program is run on a computer device, cause the computer device to execute the aforementioned fire prevention analysis method based on multi-source data fusion.

[0084] For example, computer-readable storage media can be read-only memory (ROM), random access memory (RAM), compact disc read-only memory (CD-ROM), magnetic tape, floppy disk, and optical data storage devices.

[0085] It should also be noted that, in this invention, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0086] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0087] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art. The general principles defined in the invention may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown in the invention, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A fire prevention analysis method based on multi-source data fusion, characterized in that, The specific steps include the following: S1. Collect real-time video of the monitored area, identify the dynamic evolution characteristics of smoke, analyze smoke anomalies based on the dynamic evolution characteristics of smoke, and screen suspected thermal anomaly points based on smoke anomalies. S2. Collect combustible data within a set range of suspected thermal anomalies, and analyze the flammability of combustibles based on the combustible data; S3. Analyze the combustion risk based on smoke anomalies and the flammability of combustibles at suspected thermal anomaly points; S4. Collect terrain and obstacle data within a set range of suspected thermal anomalies, and predict the oxygen entry under terrain conditions and obstacle obstruction based on the terrain and obstacle data. S5. Analyze the likelihood of a fire occurring at suspected thermal anomaly points based on the combustion risk and oxygen entry situation, and determine whether a fire warning is needed based on the likelihood of a fire occurring at suspected thermal anomaly points. S4 includes the following specific steps: S41. Collect terrain data within a defined range of suspected thermal anomalies. The terrain data includes slope and curvature. The slope is obtained using a slope calculation formula, which is: In the formula, For elevation, Let X be the rate of change of elevation in the X direction. The elevation change rate in the Y direction is given by the curvature, which is obtained using the curvature calculation formula: In the formula, Let X be the second-order elevation change rate in the X direction. Let Y be the second-order elevation change rate in the Y direction; S42. Obtain the slope influence value based on the ratio of slope to the maximum slope in the area, and obtain the curvature influence value based on the curvature normalization formula, wherein the curvature normalization formula is: In the formula, This represents the minimum curvature of the region. The topographic influence factor is obtained by weighted summation of the slope influence value and the curvature influence value to find the maximum curvature of the region. S43. Collect obstacle data within a set range of suspected thermal anomaly points, wherein the obstacle data includes obstacle area, obstacle ventilation rate, and obstacle spatial distribution coefficient; S44. Calculate the obstacle influence factor based on the obstacle influence calculation formula, wherein the obstacle influence calculation formula is: In the formula, The area of ​​the obstacle. The obstacle spatial distribution coefficient. The area of ​​the obstacle along the ventilation path. For the ventilation rate of the obstacle, This represents the proportion of air actually blocked by the obstacle. The area of ​​the ventilation path. This represents the total proportion of obstacles that prevent oxygen from entering. S45. Obtain the oxygen entry rate under terrain conditions and obstacle obstruction based on the product of the basic oxygen entry rate, terrain influence factor and obstacle influence factor.

2. The fire prevention analysis method based on multi-source data fusion according to claim 1, characterized in that, S1 includes the following specific steps: S11. Acquire real-time video of the monitored area and generate a two-dimensional motion vector for each pixel in the image sequence; S12. Extract the smoke regions from the image sequence, and obtain the average motion amplitude, motion consistency, region area growth rate, and texture change rate for each smoke region. The average motion amplitude is obtained using the average motion amplitude calculation formula, which is: In the formula, This represents the total number of pixels in the smoke area. Let be the motion vector of the i-th pixel in the smoke region. Motion consistency is obtained through the motion consistency calculation formula, which is: In the formula, Let be the direction angle of the motion vector of the i-th pixel in the smoke region. The average direction angle of all motion vectors within the smoke area is given. The area growth rate is obtained using the area growth rate calculation formula, which is: In the formula, Let be the smoke area at time t. The time interval is used to calculate the texture change rate using the texture change rate calculation formula, which is: In the formula, Let be the contrast at time t. The formula for calculating the contrast is: In the formula, Let be the joint probability of gray levels u and v; S13. Obtain the motion amplitude anomaly degree based on the ratio of the average motion amplitude to the fire smoke motion amplitude threshold, and obtain the smoke anomaly value by weighted summation based on the motion amplitude anomaly degree, motion consistency, regional area growth rate and texture change rate. S14. Based on the comparison between the smoke anomaly value and the preset smoke anomaly threshold, if the smoke anomaly value is greater than or equal to the preset smoke anomaly threshold, then the smoke area corresponding to the smoke anomaly value is set as a suspected thermal anomaly point.

3. The fire prevention analysis method based on multi-source data fusion according to claim 2, characterized in that, S2 includes the following specific steps: S21. Collect combustible data within a set range of suspected thermal anomalies, wherein the combustible data includes the ignition point, volume and porosity of the combustible. S22. Obtain combustible ignition point anomalies based on the difference between combustible ignition point and ambient temperature, obtain combustible volume anomalies based on the ratio of combustible volume to safe combustible volume threshold, and obtain porosity anomalies based on the ratio of porosity to safe porosity threshold. S23. Obtain the flammability evaluation value of the combustible material by weighted summation of the combustible material ignition point anomaly value, combustible material volume anomaly value and porosity anomaly value.

4. The fire prevention analysis method based on multi-source data fusion according to claim 3, characterized in that, S3 includes the following specific steps: The combustion risk index is obtained by multiplying the smoke anomaly value of suspected thermal anomalies with the flammability evaluation value of combustibles.

5. The fire prevention analysis method based on multi-source data fusion according to claim 4, characterized in that, S5 includes the following specific steps: S51. Obtain the probability of a fire occurring at a suspected thermal anomaly point based on the product of the combustion risk index and the oxygen entry rate. S52. Based on the comparison between the probability of a fire occurring at a suspected thermal anomaly point and a preset fire probability threshold, if the probability of a fire occurring at a suspected thermal anomaly point is greater than or equal to the preset fire probability threshold, it is determined that a fire warning needs to be issued.

6. A fire prevention analysis system based on multi-source data fusion, used to implement the fire prevention analysis method based on multi-source data fusion as described in any one of claims 1-5, characterized in that, include: The suspected thermal anomaly filtering module is used to collect real-time video of the monitored area, identify the dynamic evolution characteristics of smoke, analyze smoke anomalies based on the dynamic evolution characteristics of smoke, and filter suspected thermal anomalies based on smoke anomalies. The combustible material flammability analysis module is used to collect combustible material data within a set range of suspected thermal anomaly points and analyze the flammability of combustible materials based on the combustible material data; The combustion risk analysis module is used to analyze combustion risk based on smoke anomalies and the flammability of combustibles at suspected thermal anomaly points. The oxygen entry analysis module is used to collect terrain data and obstacle data within a set range of suspected thermal anomalies, and to predict the oxygen entry situation under terrain conditions and obstacle obstruction based on the terrain data and obstacle data. The fire early warning module is used to analyze the probability of a fire occurring at suspected thermal anomaly points based on combustion risk and oxygen entry, and to determine whether a fire early warning is needed based on the probability of a fire occurring at suspected thermal anomaly points.

7. An electronic device, characterized in that, include: A memory and a processor, wherein the memory stores a computer program that can be called by the processor, and the processor executes the fire prevention analysis method based on multi-source data fusion as described in any one of claims 1-5 by calling the computer program stored in the memory.

8. A computer-readable storage medium, characterized in that, The system stores instructions that, when executed on a computer, cause the computer to perform the fire prevention analysis method based on multi-source data fusion as described in any one of claims 1-5.

Images