A video fire spread monitoring method based on bounding box mask centroid offset
Patent Information
- Application Number
- CN202510809041.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-17
- Publication Date
- 2026-10-27
AI Technical Summary
[0003]然而，传统的火灾监测手段（如烟雾探测器、温度传感器）在工厂环境中存在诸多局限性，难以满足现代工厂复杂场景的火灾预警需求，具体问题包括覆盖范围有限，工厂通常占地面积大，且内部结构复杂，传统传感器需要密集部署才能实现全面覆盖，这不仅增加了成本，还可能导致监测盲区；误报率高，工厂环境中存在大量干扰因素，例如焊接烟雾、粉尘、蒸汽等，这些因素容易触发烟雾探测器的误报，而高温作业环境可能导致温度传感器的误报；响应速率慢，传统传感器通常需要火灾发展到一定阶段才能触发警报，难以实现火灾的早期预警，在工厂环境中，火灾往往在初期阶段发展迅速，传统传感器的延迟可能导致火灾在未被发现的情况下迅速蔓延，造成更大的损失
[0016]其有益效果在于：本发明公开了一种基于边框掩码质心偏移的视频火灾蔓延监测方法，通过部署在预设区域的多源摄像头网采集实时图像并进行去噪处理得到目标图像，对目标图像中火焰区域采用边界框进行标注得到标注图像；构建图像检测网络，并利用标注图像进行训练，然后对实时采集的视频数据进行火焰点检测，同时将检测到的火焰边界框通过最小外接矩阵映射至RGB图像中，最小外接矩阵内区域即为火焰区域；对RGB图像进行灰度及图像分割处理，将火焰区域及背景区域进行分割，然后针对分割得到的火焰区域，设计边框掩码质心偏移算法进行火焰蔓延速率及方向，从而对火灾蔓延情况进行评估。在工厂环境火灾监测任务中，本发明不仅可以精准定位火灾源头，还能实时分析火灾蔓延趋势，从而为火灾应急决策提供关键数据支持和保障。
Smart Images
Figure CN122990554A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of fire monitoring technology, and in particular to a video fire spread monitoring method based on border mask centroid offset. Background Technology
[0002] Fire refers to the damage caused by uncontrolled combustion in time or space. It is one of the most frequent and widespread major disasters threatening public safety. Factory fires, due to their unique environment and complexity, often cause even more severe consequences. They not only threaten the lives of employees but can also trigger chain reactions, such as toxic gas leaks and explosions, causing long-term impacts on surrounding communities and the ecological environment. Therefore, early detection and timely prevention of factory fires are crucial.
[0003] However, traditional fire monitoring methods (such as smoke detectors and temperature sensors) have many limitations in factory environments and cannot meet the fire early warning needs of complex modern factory scenarios. Specific problems include: limited coverage, as factories typically cover large areas and have complex internal structures, requiring dense deployment of traditional sensors to achieve comprehensive coverage, which not only increases costs but may also lead to monitoring blind spots; high false alarm rate, as there are many interfering factors in factory environments, such as welding fumes, dust, and steam, which can easily trigger false alarms from smoke detectors, while high-temperature working environments may cause false alarms from temperature sensors; and slow response speed, as traditional sensors usually require the fire to develop to a certain stage before triggering an alarm, making it difficult to achieve early fire warning. In factory environments, fires often develop rapidly in their initial stages, and the delay of traditional sensors may cause the fire to spread rapidly without being detected, resulting in greater losses.
[0004] With the expansion of industrial scale and the advancement of production technology, the demand for factory fire monitoring is increasing. Traditional technologies can no longer meet the complex needs of modern factories. Therefore, there is an urgent need for an efficient, accurate, and intelligent fire monitoring method to address the fire monitoring challenges in modern factories. Summary of the Invention
[0005] In view of this, this application provides a video fire spread monitoring method based on border mask centroid offset to address the shortcomings of the existing technology.
[0006] The first aspect of this application provides a video fire spread monitoring method based on border mask centroid offset, including: Video data is collected in real time by a multi-source camera network deployed in a preset area. Based on the video data, corresponding real-time images are obtained and noise reduction processing is performed to obtain the target image. Flame regions in all target images are annotated using flame bounding boxes to obtain an annotated image. An image detection network is constructed and trained using all labeled images. Based on the trained image detection network, flame points are detected in real-time video data, and the detected flame bounding boxes are mapped to the RGB image corresponding to the real-time image through a minimum bounding box. The area within the minimum bounding box corresponding to the flame bounding box is the flame area. The RGB image is processed to obtain a flame grayscale image; the flame region and background region in the flame grayscale image are segmented based on a preset image segmentation algorithm. Based on the flame region obtained through image segmentation, a boundary mask centroid offset algorithm is designed to calculate the flame spread rate and direction, and the first calculation result is obtained. Based on the first calculation result, the extent of fire spread is assessed.
[0007] In one possible implementation of the first aspect, acquiring the corresponding real-time image and performing denoising processing includes: The real-time image is denoised using a Gaussian filter function based on the OpenCV algorithm.
[0008] In one possible implementation of the first aspect, data annotation of the flame region in all target images using flame bounding boxes includes: Data annotation was performed using the deep learning image labeling software Labelme.
[0009] In one possible implementation of the first aspect, determining the minimum bounding box of the flame bounding box includes: Obtain all coordinates of the flame bounding box to obtain a coordinate set; Based on the coordinate set, the maximum value in the coordinate set is calculated. Coordinates are , minimum Coordinates are ,maximum Coordinates are and the smallest Coordinates are ; coordinates , , and Determine the minimum bounding matrix of the flame bounding box.
[0010] In one possible implementation of the first aspect, the design of the border mask centroid offset algorithm includes: Calculate the centroid coordinates within the flame region to obtain the second calculation result; Based on the second calculation result, the centroid coordinates of each real-time image frame within a preset time period are obtained to obtain a centroid coordinate set; Taking the centroid coordinates corresponding to the first real-time image in the centroid coordinate set as the origin, the real-time image... The axis direction is the polar coordinate axis direction. A polar coordinate system is established, and the remaining continuous centroid coordinates in the centroid coordinate set are converted into continuous polar coordinates and added to the polar coordinate system. Based on the polar coordinates of all centroid coordinates in the centroid coordinate set, the flame centroid distance offset rate and flame centroid offset angle are calculated and added to the polar coordinate system.
[0011] In one possible implementation of the first aspect, the centroid coordinates of the flame region are calculated using the following formula: coordinate Let the coordinates be the centroid coordinates of the flame region. The coordinates of the pixels in the grayscale image of the flame region in the corresponding real-time image coordinate system, and the coordinates are... The pixel value .
[0012] In one possible implementation of the first aspect, converting the remaining continuous centroid coordinates in the centroid coordinate set into continuous polar coordinates includes: Obtain the coordinates of all continuous centroids in the centroid coordinate set. , , ,……and ; With the centroid coordinates Calculate the centroid coordinates, with the origin of the polar coordinate system as the reference point. The relative coordinates with respect to the origin are as follows: The centroid coordinates The coordinates are The centroid coordinates The coordinates are ; Calculate the centroid coordinates The radius relative to the origin is specifically: Calculate the centroid coordinates The angle with respect to the origin is as follows: The centroid coordinates The corresponding polar coordinates are .
[0013] In one possible implementation of the first aspect, the calculation of the flame centroid distance offset rate using the following formula includes: For data points, For the number of data points, Corresponding to the The centroid polar radius of each data point The distance offset rate of the flame's centroid.
[0014] In one possible implementation of the first aspect, the flame centroid offset angle is calculated using the following formula: This represents the offset of the centroid angle between adjacent real-time images in the polar coordinate system. The number of frames in the real-time image. The angle of offset of the flame's center of mass.
[0015] In one possible implementation of the first aspect, assessing the degree of fire spread based on the first calculation result includes: If we determine 0 < ≤D1, the fire spread level is determined to be Level 1, and the fire spread angle is... ; If we determine that D1 < ≤D2, the fire spread level is determined to be Level II, and the fire spread angle is... ; If we determine that D2 < <1, the fire spread level is determined to be level three, and the fire spread angle is [missing information]. .
[0016] Its beneficial effects are as follows: This invention discloses a video fire spread monitoring method based on border mask centroid offset. Real-time images are acquired through a multi-source camera network deployed in a preset area, and denoising is performed to obtain the target image. The flame region in the target image is labeled with a bounding box to obtain an labeled image. An image detection network is constructed and trained using the labeled image. Then, flame point detection is performed on the real-time acquired video data. Simultaneously, the detected flame bounding boxes are mapped to the RGB image using a minimum bounding matrix; the region within the minimum bounding matrix is the flame region. The RGB image undergoes grayscale and image segmentation processing to separate the flame region from the background region. Then, for the segmented flame region, a border mask centroid offset algorithm is designed to determine the flame spread rate and direction, thereby assessing the fire spread situation. In factory environment fire monitoring tasks, this invention can not only accurately locate the fire source but also analyze the fire spread trend in real time, thus providing key data support and assurance for fire emergency decision-making. Attached Figure Description
[0017] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of this application. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.
[0018] Figure 1 This is a schematic diagram of a video fire spread monitoring method based on border mask centroid offset provided in an embodiment of this application. Detailed Implementation
[0019] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.
[0020] In this application, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, 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 limitation, 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.
[0021] Example In existing technologies, traditional fire monitoring methods have many limitations in factory environments and are difficult to meet the fire early warning needs of complex scenarios in modern factories. Specific problems include limited coverage, high false alarm rate and slow response rate. With the expansion of industrial scale and the advancement of production technology, the demand for factory fire monitoring is growing, and traditional technologies can no longer meet the complex needs of modern factories.
[0022] Therefore, this application provides a video fire spread monitoring method based on border mask centroid offset, such as... Figure 1 As shown, it includes: Video data is collected in real time by a multi-source camera network deployed in a preset area. Based on the video data, corresponding real-time images are obtained and noise reduction processing is performed to obtain the target image. Flame regions in all target images are annotated using flame bounding boxes to obtain an annotated image. An image detection network is constructed and trained using all labeled images. Based on the trained image detection network, flame points are detected in real-time video data, and the detected flame bounding boxes are mapped to the RGB image corresponding to the real-time image through a minimum bounding box. The area within the minimum bounding box corresponding to the flame bounding box is the flame area. The RGB image is processed to obtain a flame grayscale image; the flame region and background region in the flame grayscale image are segmented based on a preset image segmentation algorithm. Based on the flame region obtained through image segmentation, a boundary mask centroid offset algorithm is designed to calculate the flame spread rate and direction, and the first calculation result is obtained. Based on the first calculation result, the extent of fire spread is assessed.
[0023] The process involves real-time video data acquisition via a network of multi-source cameras deployed inside and outside the factory's flammable material storage areas. Real-time images are then obtained from this video data and denoised to produce the target image. Because the multi-source cameras continuously monitor the flammable material storage areas inside and outside the factory, resulting in high camera temperatures and dark, unevenly lit fields of view, Gaussian filtering is used to denoise the original images. Specifically, the Gaussian filtering function of the OpenCV algorithm is employed to denoise the real-time images.
[0024] The deep learning image labeling software Labelme is used to annotate the flame region in the target image with flame bounding boxes. The annotation output type is .json file, which includes the label category, the coordinates of the upper left corner of the bounding box, and the length and width information of the annotated image. The target image is then processed to obtain the annotated image.
[0025] In this process, an image detection network is constructed, and a certain number of labeled images are used as the training set to train the image detection network. The trained image detection network is then used to detect flame points in real-time acquired video data, and the detected flame bounding boxes are mapped to the acquired RGB images through the minimum bounding box matrix.
[0026] Specifically, the method for obtaining the minimum bounding matrix corresponding to any flame bounding box is as follows: obtain all coordinates of any flame bounding box to get a coordinate set; then, calculate the coordinates in this coordinate set. Maximum and minimum coordinates The maximum and minimum coordinates, then using coordinates , , and Determine the minimum bounding matrix of the flame bounding box.
[0027] In this process, the RGB image is processed to obtain a grayscale image of the flame. Then, the flame grayscale image is segmented using a preset image segmentation algorithm to distinguish the flame region from the background region. The preset image segmentation algorithm can be other algorithms such as Region Growing or Watershed, and this example does not specify a particular algorithm.
[0028] Specifically, the centroid coordinates of the flame region are calculated as follows: coordinate Let the coordinates be the centroid coordinates of the flame region. The coordinates of the pixels in the grayscale image of the flame region in the corresponding real-time image coordinate system, and the coordinates are... The pixel value .
[0029] Specifically, based on the formula for calculating the centroid coordinates of the flame region, the centroid coordinates of a video frame within 1 second are calculated. If there are n real-time images within 1 second, n consecutive centroid coordinates can be obtained. , , ,……and ; Using the centroid coordinates As the origin of the polar coordinate system, the real-time image The axis direction is the polar coordinate axis direction. A polar coordinate system is established. The subsequent centroid coordinates are added to the polar coordinate system by calculating the relative coordinates with respect to the origin. Specifically: The centroid coordinates The coordinates are The centroid coordinates The coordinates are ; Calculate the centroid coordinates The radius relative to the origin is specifically: Calculate the centroid coordinates The angle with respect to the origin is as follows: The centroid coordinates The corresponding polar coordinates are .
[0030] The calculation of the flame centroid distance offset rate using the following formula includes: For data points, For the number of data points, Corresponding to the The centroid polar radius of each data point The distance offset rate of the flame's centroid; , ; This embodiment calculates the slope, i.e., the flame centroid distance offset rate, by linearly fitting the centroid distance in polar coordinates to natural numbers. It should be noted that the inter-frame centroid distance change within 1 second can be approximated as a uniform change. In the polar coordinate system, the independent variable is assumed to be... Natural number sequence = ( =1,2,…, , (Number of data points); dependent variable : Corresponding to the The centroid radius of each point (i.e., the distance from the centroid to the origin of the polar coordinates), through a linear equation To fit the data and calculate the slope The above formula is obtained by using the least squares method for calculation and simplification.
[0031] The calculation of the flame centroid offset angle using the following formula includes: This represents the offset of the centroid angle between adjacent real-time images in the polar coordinate system. The number of frames in the real-time image. The angle of offset of the flame's center of mass.
[0032] Among them, when judging 0 < When the fire spread angle is ≤0.3, the fire spread level is determined to be Level 1 (minor spread), and the fire spread angle is... When judging 0.3 < When the fire spread angle is ≤0.7, the fire spread level is determined to be Level II (rapid spread), and the fire spread angle is... When judging D2 < When the fire spread angle is less than 1, the fire spread level is determined to be Level III (rapid spread), and the fire spread angle is... .
[0033] This embodiment includes real-time image data acquisition and denoising processing based on a multi-source camera network deployed inside and outside the factory's flammable material storage areas to obtain target images. Based on flame detection using an image detection network, the detected flame bounding boxes are mapped onto the acquired RGB images using a minimum bounding matrix to extract and visualize the flame region. An image segmentation algorithm is used to obtain a grayscale image of the flame region, and the flame region is distinguished from the background region based on the grayscale image. Finally, a designed border mask centroid offset algorithm is used to evaluate the flame spread rate and direction. This embodiment provides a video fire spread monitoring method based on border mask centroid offset to address the shortcomings of existing vision-based factory environment fire detection technologies in fire spread trend analysis, thereby providing crucial data support and technical assurance for fire emergency decision-making.
[0034] In some embodiments, acquiring a corresponding real-time image and performing noise reduction processing includes: The real-time image is denoised using a Gaussian filter function based on the OpenCV algorithm.
[0035] In some embodiments, data annotation of the flame region in all target images using flame bounding boxes includes: Data annotation was performed using the deep learning image labeling software Labelme.
[0036] In some embodiments, determining the minimum bounding box of the flame bounding box includes: Obtain all coordinates of the flame bounding box to obtain a coordinate set; Based on the coordinate set, the maximum value in the coordinate set is calculated. Coordinates are , minimum Coordinates are ,maximum Coordinates are and the smallest Coordinates are ; coordinates , , and Determine the minimum bounding matrix of the flame bounding box.
[0037] In some embodiments, the design of the border mask centroid offset algorithm includes: Calculate the centroid coordinates within the flame region to obtain the second calculation result; Based on the second calculation result, the centroid coordinates of each real-time image frame within a preset time period are obtained to obtain a centroid coordinate set; Taking the centroid coordinates corresponding to the first real-time image in the centroid coordinate set as the origin, the real-time image... The axis direction is the polar coordinate axis direction. A polar coordinate system is established, and the remaining continuous centroid coordinates in the centroid coordinate set are converted into continuous polar coordinates and added to the polar coordinate system. Based on the polar coordinates of all centroid coordinates in the centroid coordinate set, the flame centroid distance offset rate and flame centroid offset angle are calculated and added to the polar coordinate system.
[0038] In some embodiments, the centroid coordinates of the flame region are calculated using the following formula: coordinate Let the coordinates be the centroid coordinates of the flame region. The coordinates of the pixels in the grayscale image of the flame region in the corresponding real-time image coordinate system, and the coordinates are... The pixel value .
[0039] In some embodiments, converting the remaining continuous centroid coordinates in the centroid coordinate set into continuous polar coordinates includes: Obtain the coordinates of all continuous centroids in the centroid coordinate set. , , ,……and ; With the centroid coordinates Calculate the centroid coordinates, with the origin of the polar coordinate system as the reference point. The relative coordinates with respect to the origin are as follows: The centroid coordinates The coordinates are The centroid coordinates The coordinates are ; Calculate the centroid coordinates The radius relative to the origin is specifically: Calculate the centroid coordinates The angle with respect to the origin is as follows: The centroid coordinates The corresponding polar coordinates are .
[0040] In some embodiments, the flame centroid distance offset rate is calculated using the following formula: For data points, For the number of data points, Corresponding to the The centroid polar radius of each data point The distance offset rate of the flame's centroid.
[0041] In some embodiments, the flame centroid offset angle is calculated using the following formula: This represents the offset of the centroid angle between adjacent real-time images in the polar coordinate system. The number of frames in the real-time image. The angle of offset of the flame's center of mass.
[0042] In some embodiments, assessing the degree of fire spread based on the first calculation result includes: If we determine 0 < ≤D1, the fire spread level is determined to be Level 1, and the fire spread angle is... ; If we determine that D1 < ≤D2, the fire spread level is determined to be Level II, and the fire spread angle is... ; If we determine that D2 < <1, the fire spread level is determined to be level three, and the fire spread angle is [missing information]. .
[0043] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computing software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.
[0044] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.
[0045] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.
Claims
1. A video fire spread monitoring method based on border mask centroid offset, characterized in that, include: Video data is collected in real time by a multi-source camera network deployed in a preset area. Based on the video data, corresponding real-time images are obtained and noise reduction processing is performed to obtain the target image. Flame regions in all target images are annotated using flame bounding boxes to obtain an annotated image. An image detection network is constructed and trained using all labeled images. Based on the trained image detection network, flame points are detected in real-time video data, and the detected flame bounding boxes are mapped to the RGB image corresponding to the real-time image through a minimum bounding box. The area within the minimum bounding box corresponding to the flame bounding box is the flame area. The RGB image is processed to obtain a grayscale image of the flame; Based on a preset image segmentation algorithm, the flame region and background region in the flame grayscale image are segmented; Based on the flame region obtained through image segmentation, a boundary mask centroid offset algorithm is designed to calculate the flame spread rate and direction, and the first calculation result is obtained. Based on the first calculation result, the extent of fire spread is assessed.
2. The video fire spread monitoring method based on border mask centroid offset according to claim 1, characterized in that, Acquiring and denoising the corresponding real-time images includes: The real-time image is denoised using a Gaussian filter function based on the OpenCV algorithm.
3. The video fire spread monitoring method based on border mask centroid offset according to claim 1, characterized in that, Flame regions in all target images are annotated using flame bounding boxes, including: Data annotation was performed using the deep learning image labeling software Labelme.
4. The video fire spread monitoring method based on border mask centroid offset according to claim 1, characterized in that, Determining the minimum bounding box of the flame bounding box includes: Obtain all coordinates of the flame bounding box to obtain a coordinate set; Based on the coordinate set, the maximum value in the coordinate set is calculated. Coordinates are , minimum Coordinates are ,maximum Coordinates are and the smallest Coordinates are ; coordinates , , and Determine the minimum bounding matrix of the flame bounding box.
5. The video fire spread monitoring method based on border mask centroid offset according to claim 1, characterized in that, The design of the border mask centroid offset algorithm includes: Calculate the centroid coordinates within the flame region to obtain the second calculation result; Based on the second calculation result, the centroid coordinates of each real-time image frame within a preset time period are obtained to obtain a centroid coordinate set; Taking the centroid coordinates corresponding to the first real-time image in the centroid coordinate set as the origin, the real-time image... The axis direction is the polar coordinate axis direction. A polar coordinate system is established, and the remaining continuous centroid coordinates in the centroid coordinate set are converted into continuous polar coordinates and added to the polar coordinate system. Based on the polar coordinates of all centroid coordinates in the centroid coordinate set, the flame centroid distance offset rate and flame centroid offset angle are calculated and added to the polar coordinate system.
6. The video fire spread monitoring method based on border mask centroid offset according to claim 5, characterized in that, The centroid coordinates of the flame region are calculated using the following formula: coordinate Let the coordinates be the centroid coordinates of the flame region. The coordinates of the pixels in the grayscale image of the flame region in the corresponding real-time image coordinate system, and the coordinates are... The pixel value .
7. The video fire spread monitoring method based on border mask centroid offset according to claim 5, characterized in that, Converting the remaining continuous centroid coordinates in the centroid coordinate set to continuous polar coordinates includes: Obtain the coordinates of all continuous centroids in the centroid coordinate set. , , ,……and ; With the centroid coordinates Calculate the centroid coordinates, with the origin of the polar coordinate system as the reference point. The relative coordinates with respect to the origin are as follows: The centroid coordinates The coordinates are The centroid coordinates The coordinates are ; Calculate the centroid coordinates The radius relative to the origin is specifically: Calculate the centroid coordinates The angle with respect to the origin is as follows: The centroid coordinates The corresponding polar coordinates are .
8. The video fire spread monitoring method based on border mask centroid offset according to claim 5, characterized in that, The flame centroid distance offset rate is calculated using the following formula: For data points, For the number of data points, Corresponding to the The centroid polar radius of each data point The distance offset rate of the flame's centroid.
9. A video fire spread monitoring method based on border mask centroid offset according to claim 8, characterized in that, The flame centroid offset angle is calculated using the following formula: This represents the offset of the centroid angle between adjacent real-time images in the polar coordinate system. The number of frames in the real-time image. The angle of offset of the flame's center of mass.
10. A video fire spread monitoring method based on border mask centroid offset according to claim 9, characterized in that, Based on the first calculation result, the assessment of the fire spread includes: If we determine 0 < ≤D1, the fire spread level is determined to be Level 1, and the fire spread angle is... ; If we determine that D1 < ≤D2, the fire spread level is determined to be Level II, and the fire spread angle is... ; If we determine that D2 < <1, the fire spread level is determined to be level three, and the fire spread angle is [missing information]. .