Digital visual management method and system for security and protection equipment based on Internet of Things

Abstract

The invention relates to the technical field of image recognition, in particular to a digital visual management method and system for security and protection equipment based on the Internet of Things. The method comprises the following steps: collecting a security area image set, and carrying out road image area segmentation to obtain a fire fighting access image set and a traffic road image set; identifying a traffic obstruction road image set in the traffic road image set according to the traffic road image set; comparing road width difference values of intersection roads between the fire fighting access image set and the traffic obstruction road image set, and identifying a fire fighting vehicle traffic obstruction area; identifying communication obstacles and obstacles to be cleaned on the fire-fighting vehicle passing obstruction area; and according to the security and protection area image set, identifying a spatial free area, allocating placement areas of the spatial free area to the communicable obstacles and the to-be-cleaned obstacles to obtain an obstacle placement strategy, and uploading the obstacle placement strategy to the Internet of Things. According to the invention, the automation and response efficiency of urban fire fighting access management and control can be improved.

Description

Technical Field

[0001] This invention relates to the field of image recognition technology, and in particular to a digital visualization management method and system for security equipment based on the Internet of Things. Background Technology

[0002] In the urban public safety management system, fire lanes serve as emergency rescue routes, and their unobstructed access directly impacts the efficiency of personnel evacuation and emergency rescue during emergencies such as fires. However, due to outdated management methods and insufficient monitoring coverage, the illegal occupation and blockage of fire lanes remain widespread, commonly found in residential areas, commercial districts, and logistics zones within industrial parks. Illegal parking, temporary storage of debris, and construction equipment obstructing access cause fire trucks to encounter problems such as insufficient turning radius and obstructed straight-line passage during emergency response, seriously threatening public safety.

[0003] Traditional security management methods largely rely on manual inspections of fixed-point cameras or timed photo recognition, lacking the ability to perceive dynamic scenes in real time and understand spatial semantics. They cannot comprehensively perceive changes in the status of fire lanes and lack structured analysis methods for the type, location, and degree of obstruction. Furthermore, existing monitoring systems generally suffer from isolated sensing data, uncontrollable equipment information, and invisible task feedback, leading to delayed management response, unclear dispatch paths, and significant difficulties in tracing responsibility. Summary of the Invention

[0004] Therefore, the present invention needs to provide a digital visualization management method and system for security equipment based on the Internet of Things to solve at least one of the above-mentioned technical problems.

[0005] To achieve the above objectives, a digital visualization management method for security equipment based on the Internet of Things (IoT) includes the following steps:

[0006] Step S1: Collect a set of security area images, segment the security area image set into road image regions, and obtain a set of fire lane images and a set of traffic road images;

[0007] Step S2: Calculate the spatial gap between obstacles and road edges on the road based on the road image set to obtain road gap data; identify the road obstruction image set in the road image set based on the road gap data;

[0008] Step S3: Compare the road width difference between the intersection of the fire lane image set and the road obstruction image set, and identify the fire truck passage obstruction area based on the road width difference. The fire truck passage obstruction area includes the fire truck turning obstruction area and the fire truck straight passage obstruction area.

[0009] Step S4: Use the image set of the passageway to identify the characteristics of obstacles in the area obstructing the passage of fire trucks, and classify the obstacles into communicable obstacles and obstacles to be cleared according to the characteristics of the obstacles;

[0010] Step S5: Identify vacant areas based on the security area image set, allocate placement areas for communicable obstacles and obstacles to be cleared, obtain an obstacle placement strategy, and upload the obstacle placement strategy to the Internet of Things to perform map visualization tasks.

[0011] Optionally, step S1 includes:

[0012] Step S11: Collect image sets of the security area through the pre-deployed monitoring equipment in the security area;

[0013] Step S12: Extract low-brightness connected regions from the security area image and merge small occluded areas within the low-brightness connected regions to obtain the initial road outline map;

[0014] Step S13: Identify areas in the initial road outline map that contain red lines, yellow diagonal grid lines, and fire lane text signs as candidate fire lane image segments, and use the remaining areas in the initial road outline map as traffic road image segments;

[0015] Step S14: Construct a local polar coordinate system with the installation point of the monitoring equipment as the origin;

[0016] Step S15: Using the local polar coordinate system, the angles of the candidate fire lane image segments and the traffic road image segments are calibrated and their positions are recorded to obtain the fire lane image set and the traffic road image set.

[0017] Optionally, step S12 includes:

[0018] Step S121: Extract the brightness channel of each frame image in the security area image set, set the brightness threshold to 90, and extract the pixels with brightness below 90 in each frame image to form a low brightness pixel mask image.

[0019] Step S122: Identify continuous low-brightness regions in the low-brightness pixel mask image and filter out isolated regions with an area of ​​less than 200 pixels to obtain candidate road segments.

[0020] Step S123: Set the maximum interpolation radius to 10 pixels, perform pixel value estimation on the areas with obvious gray-level changes in the candidate road segments, and use the pixel value estimation results to fill in the areas with obvious gray-level changes to obtain a low-brightness connected region map;

[0021] Step S124: Extract the image contour edges of the low-brightness connected region map, and set the edge gradient magnitude of the image contour edges to the edge closure threshold to extract the closed road contour and construct the initial road contour map.

[0022] Optionally, calculating the spatial clearance between obstacles and the road edge in step S2 includes:

[0023] Set the frame difference threshold to 20, perform difference on the time series of consecutive frames of the road image set to obtain the inter-frame difference, and mark the pixels with an inter-frame difference greater than 20 as the foreground region;

[0024] The foreground region mask is calculated based on the inter-frame difference and color difference of the foreground region, thereby identifying dynamic obstacle regions;

[0025] Identify boundary contours in dynamic obstacle regions with dimensions between 300 and 5000 pixels and aspect ratios between 0.5 and 2.5, and record the center coordinates, side length, and orientation angle of the smallest bounding rectangle of the boundary contour to construct an obstacle structure index table;

[0026] The boundary lines on both sides of the road are determined based on the gray-level gradient changes in the upper and lower halves of each image in the road image set.

[0027] The shortest Euclidean distance from the center coordinates of the obstacle structure index table to the boundary lines on both sides of the road is calculated, and the shortest Euclidean distance is normalized with the overall road width in the traffic road image set to obtain the traffic road gap data.

[0028] Optionally, identifying dynamic obstacle regions includes:

[0029] Calculate the differences of the foreground region in the R, G, and B channels in adjacent time frames to obtain the RGB channel differences;

[0030] Extract connected regions from the foreground region where the sum of the RGB channel differences is greater than 50, and construct a color change response map of the connected region.

[0031] The foreground region mask is obtained by pixel-by-pixel superimposing of the inter-frame difference and the color change response map;

[0032] Morphological opening is performed on the foreground region mask to remove scale noise interference regions, and connected component extraction is performed on the continuous regions in the foreground region mask after removing scale noise interference regions to obtain the dynamic obstacle region.

[0033] Optionally, step S3 includes:

[0034] Step S31: Align the fire lane image set with the congested road image set by coordinates, extract the spatially overlapping area, and form a subset of intersecting road images;

[0035] Step S32: Extract the road boundary lines from the subset of intersecting road images, and perform perspective distortion correction on the road boundary lines based on the camera intrinsics and shooting height of the monitoring equipment to obtain the actual intersecting road images;

[0036] Step S33: Measure the pixel width of each road outline in the actual intersection road image and convert it into the real scene width to obtain the actual width of the intersection roads;

[0037] Step S34: Extract the standard fire lane width from the fire lane image set, compare it with the actual width of the intersecting roads, calculate the road width difference, and set the width deviation judgment threshold to 0.6 meters. Mark all image areas in the intersecting road image subset where the road width difference is greater than 0.6 meters as suspected obstruction areas.

[0038] Step S35: Determine the type of obstruction for suspected obstruction areas to obtain the areas obstructing the passage of fire trucks.

[0039] Optionally, step S35 includes:

[0040] Step S351: Determine the main direction of the suspected obstruction region in the local polar coordinate system based on the angle pointed to by the long side of the suspected obstruction region, and obtain the main direction of the obstruction region;

[0041] Step S352: Use the image principal axis of the intersection road image of the obstruction area to determine the type. If the angle between the main direction of the obstruction area and the image principal axis is less than or equal to 45 degrees, the suspected obstruction area is determined to be a straight obstruction area for fire trucks; if the angle between the main direction of the obstruction area and the image principal axis is greater than 45 degrees, the suspected obstruction area is determined to be a candidate area for turning obstruction.

[0042] Step S353: Obtain the minimum turning radius of the fire truck through the Internet of Things, and determine the turning accessibility based on the obstacle boundary features in the actual intersection road image corresponding to the candidate turning obstacle area and the minimum turning radius of the fire truck. If the shortest distance from the center of the inner corner of the road to the nearest obstacle boundary in the corresponding actual intersection road image is less than the minimum turning radius of the fire truck, then the candidate turning obstacle area is determined to be the turning obstacle area of ​​the fire truck.

[0043] Step S354: Spatially merge the fire truck straight-through obstruction area and the fire truck turning obstruction area to obtain the fire truck passage obstruction area.

[0044] Optionally, the obstacle classification in step S4 includes:

[0045] Extract the spatial location, bounding rectangle size, angle direction, duration in frames, and motion speed estimate of all obstacles in the area obstructing the passage of fire trucks, and construct a dynamic feature matrix of obstacles;

[0046] Extract image texture features corresponding to the areas obstructing the passage of fire trucks, identify high contour regularity areas and low contour regularity areas based on image texture features, and filter candidate obstacles for communication devices from the high contour regularity areas by combining the obstacle dynamic feature matrix.

[0047] Identify the image and text features corresponding to candidate obstacles that can communicate with devices, and access the IoT device database for text index matching. If a matching device text index exists, mark the candidate obstacle that can communicate with devices as a communication obstacle.

[0048] The relative three-dimensional center of the remaining obstacles that are not marked as communicable obstacles in the area obstructing the passage of fire trucks is extracted, and the straight-line distance between the relative three-dimensional center of the remaining obstacles and the main traffic axis in the actual intersection road image is calculated. If the straight-line distance is less than 1.2 meters and the remaining obstacle belongs to a low contour regularity area, the remaining obstacle is determined to be an obstacle to be cleared.

[0049] Optionally, step S5 includes:

[0050] Step S51: Extract the texture continuity and grayscale balance of the ground plane view area in each image of the security area image set, and identify candidate spatial empty areas;

[0051] Step S52: Perform spatial unitization processing on the candidate empty space area, remove spatial units smaller than the set minimum placement area threshold of 1.5 square decimeters, and calculate the centroid coordinates, the side length of the maximum inscribed rectangle and the accessibility of adjacent channels for each remaining empty unit to generate a structured spatial area table.

[0052] Step S53: Use the structured spatial region table to perform spatial shape adaptation matching on the communicable obstacles, and add the communication contact number in the device text index to the adaptation matching to obtain the communicable obstacle placement strategy.

[0053] Step S54: Based on the structured spatial region table, minimize the center-of-gravity distance of the obstacles to be cleared to obtain the obstacle placement strategy;

[0054] Step S55: Spatially merge the communicable obstacle placement strategy with the obstacle placement strategy to be cleared to obtain the obstacle placement strategy, and upload the obstacle placement strategy to the Internet of Things to perform the map visualization task.

[0055] This invention introduces multiple steps, including image spatial reconstruction, structural region extraction, obstacle type identification, intelligent device matching, and optimized placement allocation, forming a closed-loop management system that integrates image semantic analysis and IoT device interaction. This system can systematically solve core problems in traditional security management, such as passive monitoring, delayed response, and data fragmentation. In terms of image perspective processing, perspective correction is performed on each frame to convert the original tilted or upward-looking images into approximately top-down orthographic projection images. This operation is achieved by constructing a homography matrix based on camera installation parameters (such as focal length, pitch angle, and height), providing spatial geometric reconstruction capabilities and accurately restoring the spatial layout of ground structures. The orthographic projection image ensures that the area, positional relationships, and orientation angles of subsequently measured regions are more consistent with actual proportions, thus providing a high-precision foundation for obstacle detection and spatial assessment. During region scanning, a fixed-size sliding window is used to analyze the image block by block. The window size is set to 32×32 pixels, ensuring a balance between granularity and efficiency in feature detection. Texture continuity metrics (based on gray-level co-occurrence matrix evaluation) are extracted from each window, with a continuity threshold of 0.75 to ensure that identified candidate empty regions possess stable material structures. For dynamic obstacle recognition, the foreground region's RGB channel difference changes are analyzed frame-by-frame, and the sum of changes across the three channels is calculated. Only pixels with a sum greater than 50 are processed to avoid misjudgments caused by minor lighting variations. Subsequently, a morphological opening operation with a 5×5 structuring element is used to remove discrete noise regions, ensuring complete obstacle contours and clear boundaries. Furthermore, by extracting connected regions and analyzing their change rate, duration (number of frames), size, and orientation angle, a dynamic obstacle feature matrix is ​​constructed to support subsequent obstacle type recognition. During type determination, the obstacle's principal direction is determined based on the direction of its circumscribed rectangle's long side, while the image's principal axis is obtained by analyzing the direction of the road edge lines in the image, aligning with the actual fire truck's travel direction. When the angle between the obstacle's principal direction and the image's principal axis is less than or equal to 45°, it primarily obstructs straight-line travel; if the angle is greater than 45°, it may affect turning operations. The 45° threshold set here clearly distinguishes the types of obstacles affecting different traffic intentions, contributing to the logical clarity of path determination. Combining fire truck structural parameters obtained from the IoT platform, especially the minimum turning radius (e.g., set to 5.8 meters), it determines whether obstacles will obstruct the corner path, improving the physical matching accuracy of the traffic path. For obstacle classification and recognition, an image texture regularity recognition mechanism is introduced, classifying obstacles with clear outlines and neat geometric boundaries as high-regularity areas, and further filtering out potential communication devices. By recognizing textual information on the device's appearance (such as numbers, labels, etc.) and matching it with the index structure in the IoT device database, a successful match is identified as a communicable obstacle.This process not only bridges the gap between physical identification and digital device identity but also supports subsequent response operations such as remote control and call notification based on device number. For other obstacles that cannot be identified through communication, their three-dimensional center point is extracted, and the nearest perpendicular distance to the image's principal axis is calculated. If this distance is less than 1.2 meters, it indicates that the obstacle is located on the main passageway and has a high degree of obstruction to traffic, thus being identified as an obstacle to be cleared. The 1.2-meter threshold is based on the reserved space design for the width of common fire lane vehicles, ensuring the practicality of the judgment. In terms of spatial placement planning, the minimum unit processing is performed based on the identified candidate empty areas. Areas smaller than 1.5 square decimeters in the empty areas are eliminated. This area threshold is set according to the minimum placement requirements of common mobile equipment (such as fire extinguisher boxes, wheeled equipment, etc.) to avoid misjudging fragmented areas and improve resource utilization efficiency. Subsequently, the system matches suitable spatial units to communicable obstacles and adds a contact information field for the equipment to the placement strategy, facilitating on-site personnel to contact and handle the situation. For obstacles to be cleared, a strategy of minimizing center-of-gravity distance is used to assign them to the nearest compliant vacant area, optimizing the relocation route and reducing manual intervention time and path length. Finally, the resulting placement strategy table is uploaded to the IoT management platform, and in conjunction with the map visualization system, the obstacle type, location information, placement plan, and corresponding handling suggestions are visualized. This achieves closed-loop digital management of the entire process from perception, identification, judgment to disposal, greatly improving the automation, visualization, and response efficiency of urban fire lane control.

[0056] Optionally, this specification also provides an IoT-based digital visualization management system for security equipment, used to execute the IoT-based digital visualization management method for security equipment as described above. The IoT-based digital visualization management system for security equipment includes:

[0057] The image segmentation module is used to acquire a set of security area images, segment the security area image set into road image regions, and obtain a set of fire lane images and a set of traffic road images;

[0058] The obstruction road recognition module is used to calculate the spatial gap between obstacles and road edges on the road based on the road image set, and obtain road gap data; and to identify the obstruction road image set in the road image set based on the road gap data.

[0059] The fire truck access obstruction area identification module is used to compare the road width difference between the intersection road between the fire lane image set and the access obstruction road image set, and identify the fire truck access obstruction area based on the road width difference. The fire truck access obstruction area includes the fire truck turning obstruction area and the fire truck straight-through obstruction area.

[0060] The obstacle classification module is used to identify the characteristics of obstacles in the area where fire trucks are obstructing passage using a set of images of the passing roads, and classify the obstacles into communicable obstacles and obstacles to be cleared based on the characteristics of the obstacles.

[0061] The obstacle placement module is used to identify vacant areas based on the security area image set, allocate placement areas of vacant areas for communicable obstacles and obstacles to be cleared, obtain obstacle placement strategies, and upload the obstacle placement strategies to the Internet of Things to perform map visualization tasks.

[0062] The present invention relates to an IoT-based digital visualization management system for security equipment. This system can implement any of the IoT-based digital visualization management methods of the present invention. It is used to connect the operation and signal transmission media between various modules to complete the IoT-based digital visualization management method for security equipment. The internal modules of the system cooperate with each other, thereby improving the automation, visualization and response efficiency of urban fire lane control. Attached Figure Description

[0063] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0064] Figure 1 This is a flowchart illustrating the steps of the Internet of Things-based digital visualization management method for security equipment according to the present invention.

[0065] Figure 2 This is a detailed flowchart of step S1 in the present invention;

[0066] Figure 3 This is a schematic diagram illustrating the application scenario of step S3 in this invention;

[0067] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0068] The technical method of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0069] Furthermore, the accompanying drawings are merely illustrative of the invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor methods and / or microcontroller methods.

[0070] It should be understood that although the terms "first," "second," etc., may be used herein to describe various units, these units should not be limited by these terms. These terms are used merely to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, a first unit may be referred to as a second unit, and similarly, a second unit may be referred to as a first unit. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0071] To achieve the above objectives, please refer to Figures 1 to 3 This invention provides a digital visualization management method for security equipment based on the Internet of Things, the method comprising the following steps:

[0072] Step S1: Collect a set of security area images, segment the security area image set into road image regions, and obtain a set of fire lane images and a set of traffic road images;

[0073] In this embodiment, a high-density urban residential area was selected as the target security area. Twelve fixed-point cameras, all high-definition wide-angle cameras, were deployed within the area, covering key areas such as entrances / exits, main passageways, building entrances, and turning sections. All devices were set to a 30° overhead angle, a height of 2.5 meters, and an image resolution of 1920×1080, with a frame rate of 10 frames per minute. The image sets were processed hourly. First, road and non-road areas were divided based on color zoning and texture feature differences. Fire lanes were extracted based on paving material (dark gray asphalt) and linear boundaries; other traffic roads within the block were identified as auxiliary passageways based on geometric connectivity and edge integrity. After initial segmentation, two types of image sets were obtained: one set consisted of fire lane images with clear traffic guidance lines; the other set consisted of unmarked but open traffic road images, used for subsequent obstacle recognition and spatial matching.

[0074] Step S2: Calculate the spatial gap between obstacles and road edges on the road based on the road image set to obtain road gap data; identify the road obstruction image set in the road image set based on the road gap data;

[0075] In this embodiment, the spatial gap between obstacles and road boundaries in each image of a passable road is analyzed. The boundary is extracted based on the gray-scale inflection points on both sides of the image, and its accuracy is calibrated by referring to the spacing of ground markings, with a conversion ratio of 0.022 meters per pixel. The vertical distance between the shortest distance points of several obstacles and the road edge is collected within the image (one set of measurement points is taken every 5 frames), forming an "obstacle gap data table". Taking a certain image as an example, the closest distance between the right wheel of a vehicle and the road edge is 0.38 meters, which is less than the set safety threshold of 0.5 meters. Therefore, this image is labeled as "edge obstruction". Multiple labeled images form clustered areas in the spatial distribution map, which are then marked as "obstructed traffic" road sections, forming a set of obstructed road images to provide a basis for subsequent comparison.

[0076] Step S3: Compare the road width difference between the intersection of the fire lane image set and the road obstruction image set, and identify the fire truck passage obstruction area based on the road width difference. The fire truck passage obstruction area includes the fire truck turning obstruction area and the fire truck straight passage obstruction area.

[0077] In this embodiment, to further confirm whether there are serious obstructions in the fire lane, the identified "fire lane images" and "obstructed passage images" are spatially overlapped and matched to extract image groups taken in the same physical area. Image alignment is performed using camera number, geographic coordinates, and shooting angle as registration criteria, and time tags are used to select images taken at similar times. For overlapping area images, double-edge extraction is used to form the lines on both sides of the road. The horizontal width is measured every 50 pixels at the image center, accumulating 5-10 sample width values, which are then converted to the real-world distance using the camera's ground calibration ratio. In one area, the measured image width was only 2.4 meters, while the standard fire lane is 3.5 meters, exceeding the critical deviation of 0.6 meters. This area was recorded as a "suspected fire truck obstruction area." Further analysis is performed to determine the angle between the image outline direction and the fire truck's direction of travel. If the directions are consistent (angle less than 20°), it is classified as "straight-through obstruction"; if it is near 90°, it is recorded as "turning obstruction."

[0078] Step S4: Use the image set of the passageway to identify the characteristics of obstacles in the area obstructing the passage of fire trucks, and classify the obstacles into communicable obstacles and obstacles to be cleared according to the characteristics of the obstacles;

[0079] In this embodiment, for identified obstruction areas, the color, outline, and changes of obstacles in multiple frames of images are analyzed to construct an object movement state table. If the object outline remains stationary for 10 consecutive frames (position change less than 5 pixels) and has vehicle texture or obvious box edge features, it is identified as a static obstacle. Further, details such as whether a house number, a contact card, or a sign placed inside the vehicle window are visible in the image are used to determine whether it is communicable. If a text area is extracted from the image (character boundary recognition success rate is higher than 85%) and a number or QR code is successfully extracted, the obstacle is marked as a "communicable obstacle"; if it is debris, discarded materials, or a vehicle without identification labels, it is classified as an "obstacle to be cleared". For example, in the image set, a white van parked for 12 hours without moving is identified in front of a residential building, with a contact number affixed to its window, and is classified as "communicable"; while in another location, a discarded sofa is piled up by the passageway without any markings or contact information, and is classified as "to be cleared".

[0080] Step S5: Identify vacant areas based on the security area image set, allocate placement areas for communicable obstacles and obstacles to be cleared, obtain an obstacle placement strategy, and upload the obstacle placement strategy to the Internet of Things to perform map visualization tasks.

[0081] In this embodiment, region reconstruction is performed on all images to identify ground areas in the security images that are free from object occlusion, have high grayscale uniformity, and continuous texture, as candidate empty spaces. These empty areas are divided into several spatial cells, each with an area calculation precision of 0.1 square meters; areas smaller than a set threshold (0.15 square meters) are discarded. The geometric centroid, bounding box size, and distance to adjacent channels are calculated for the remaining cells to form a structured spatial resource table. Next, two spatial allocation strategies are applied to the classified obstacles: communicable obstacles are preferentially matched to empty areas with the closest outline size, and their placement points are set based on the location information field in their communication numbers; obstacles to be cleared are relocated based on the shortest distance from their centroid to an empty area. Finally, all strategies are output to the security IoT platform, transforming them into an interactive visual map. The map layers are marked with the original location of the obstacle, suggested placement points, and corresponding contact information, allowing community property management and security personnel to coordinate execution and issue warnings.

[0082] Optionally, step S1 includes:

[0083] Step S11: Collect image sets of the security area through the pre-deployed monitoring equipment in the security area;

[0084] In this embodiment, 16 sets of fixed monitoring equipment, model IPC-B423-H, were pre-deployed in a security area of ​​an urban residential community covering an area of ​​approximately 25,000 square meters. These devices support automatic infrared illumination, wide-angle imaging, and low-light enhancement. Each set of equipment is installed on a 3-meter-high light pole, with an orientation set at an angle of approximately 25° to the ground, covering an average width of 16 meters. The acquisition tasks are uniformly scheduled through an IoT platform, acquiring and storing one high-quality image frame every 10 minutes to generate an image set for subsequent processing. The image acquisition period is mainly concentrated between 2:00 AM and 4:00 AM, as there are fewer dynamic obstructions during this time, facilitating subsequent structure extraction. Simultaneously, the acquisition time, device number, and physical orientation of each image frame are recorded during the acquisition process, laying the foundation for subsequent spatial correction.

[0085] Step S12: Extract low-brightness connected regions from the security area image and merge small occluded areas within the low-brightness connected regions to obtain the initial road outline map;

[0086] In this embodiment, the overall brightness distribution of the image is partitioned, and connected regions with an average brightness below 45 (based on an 8-bit grayscale image) are extracted as suspicious road contours. These regions typically correspond to asphalt or cement pavement, exhibiting strong brightness consistency. If small, isolated bright areas (less than 600 pixels²) are found within a connected region, they are considered interference caused by reflections, fallen leaves, or obstacle shadows, and are bridged through neighborhood expansion. After processing, a continuous, closed initial road contour image is obtained, and its edges are smoothed by 5 pixels to eliminate false contours. In practical applications, if the image shooting angle deviation exceeds 30°, the processing accuracy decreases by approximately 8%. Therefore, image validity needs to be determined beforehand, filtering out image frames with shooting quality below a threshold (image contrast below 10%).

[0087] Step S13: Identify areas in the initial road outline map that contain red lines, yellow diagonal grid lines, and fire lane text signs as candidate fire lane image segments, and use the remaining areas in the initial road outline map as traffic road image segments;

[0088] In this embodiment, regions with specific marking attributes are identified using color distribution features and spatial texture shape in the obtained road outline map. Red straight lines in the image (hue within the range of [350°, 10°], saturation greater than 0.6) are identified as fire lane markings, while yellow diagonal lines are determined as no-parking grid areas based on the intersection spacing and texture direction. Furthermore, text areas on the ground are identified through contour detection, and confirmed by combining character color, arrangement direction, and common keywords (such as the four characters "fire lane"). Image segments containing the above features are designated as candidate fire lane image segments, while other road outline portions not containing any of the above features are classified as general traffic road image segments. If multiple adjacent areas have the same marking features during this process, they are merged into a single complete segment to preserve the continuity of the passageway.

[0089] Step S14: Construct a local polar coordinate system with the installation point of the monitoring equipment as the origin;

[0090] In this embodiment, after acquiring image fragments of fire lanes and access roads, a local polar coordinate structure is established, using the fixed installation point of each monitoring device as the local coordinate origin, to unify subsequent spatial comparison and matching analysis. This coordinate structure uses the direction directly in front of the monitoring device's lens as the 0-degree polar axis, dividing the space into 0–360° angles clockwise. The radial unit is based on image pixels, converted to real-world metric coordinates using device calibration parameters (average conversion ratio is 1 pixel = 0.025 meters). This structure facilitates angle annotation and relative distance positioning of any contour, obstacle, or passage fragment in the image. For example, in image number A07, the identified candidate fire lane fragment is located 20 degrees to the left of the camera's orientation, with a radial distance of approximately 5.5 meters, recorded as (340°, 5.5m), providing a basis for subsequent cross-device area comparison.

[0091] Step S15: Using the local polar coordinate system, the angles of the candidate fire lane image segments and the traffic road image segments are calibrated and their positions are recorded to obtain the fire lane image set and the traffic road image set.

[0092] In this embodiment, after extracting the spatial information of image segments, key indicators such as angle, radial length, and contour area in polar coordinates are recorded for each candidate segment. For fire lane image segments, the presence of salient elements such as red lines and text markings is additionally marked; if these are missing, their confidence level is set to medium. The spatial record of the access road image segments includes the start and end angle segments, the range of the angle with the origin, and the contour direction. The information of all image segments is uniformly formatted and stored in the regional image database, and bound to the monitoring device number, shooting time, and frame number to form a structured image set. Among them, the fire lane image set is used for subsequent obstruction comparison and passage width assessment, while the access road image set is used for obstacle detection and spatial distribution analysis, forming the basic spatial model of the access structure of the security area.

[0093] Optionally, step S12 includes:

[0094] Step S121: Extract the brightness channel of each frame image in the security area image set, set the brightness threshold to 90, and extract the pixels with brightness below 90 in each frame image to form a low brightness pixel mask image.

[0095] In this embodiment, a security area image set of a residential community is selected. Each frame of the image has a resolution of 1920×1080 pixels and uses YUV format. In the preprocessing stage, the Y channel is extracted as the brightness data source. For the Y channel in each frame of the image, the brightness threshold is set to 90 (based on the grayscale range of 0-255). Pixels with values ​​lower than this value are considered as possible ground areas, forming a single-channel mask image. All pixels that meet the conditions are assigned a value of 1, and the rest are assigned a value of 0, resulting in a binary low-brightness pixel mask image. This image is used to extract potential road base areas in the subsequent process. The brightness threshold setting is obtained through statistical analysis of 400 sets of sample images. It is suitable for most outdoor daytime-cloudy shooting scenarios and can be dynamically adjusted to adapt to images acquired under nighttime supplementary lighting conditions.

[0096] Step S122: Identify continuous low-brightness regions in the low-brightness pixel mask image and filter out isolated regions with an area of ​​less than 200 pixels to obtain candidate road segments.

[0097] In this embodiment, after reading the low-brightness pixel mask image, the area of ​​each connected region is judged, and the connected structure is marked by an 8-neighborhood connection method. The total number of pixels of each connected region is calculated. Isolated regions with an area of ​​less than 200 pixels are considered to be local occlusions or abnormal reflections caused by non-road elements such as image noise, tree shadows, stains, and bird droppings, and are directly removed and not involved in subsequent processing. The remaining connected regions are retained as candidate road segments. The region boundary information is extracted by a rectangle and the position and outer width and height values ​​of the region in the image coordinates are marked. The current image frame number and acquisition time information are attached during recording for subsequent time matching. In the actual measured cell sample, an average of 4 to 7 candidate road segments can be extracted from each image.

[0098] Step S123: Set the maximum interpolation radius to 10 pixels, perform pixel value estimation on the areas with obvious gray-level changes in the candidate road segments, and use the pixel value estimation results to fill in the areas with obvious gray-level changes to obtain a low-brightness connected region map;

[0099] In this embodiment, the grayscale continuity characteristics of the obtained candidate road segment regions are further analyzed, with particular attention paid to grayscale abrupt change areas within the region, such as water stain reflections, man-made damage, and projections of temporary small obstructions. A maximum pixel interpolation radius of 10 pixels is set at these locations as the effective spatial range of the surrounding reference area. Within this radius, the pixel values ​​of the missing areas are estimated and compensated. This process does not involve redrawing the entire image; only local strong change points within the candidate road segment are processed at the pixel level. The estimation method is based on the grayscale mean of surrounding pixels and the trend of continuous directional changes, generating new pixel values ​​to fill in the abnormal points. This results in a low-brightness connected region map with uniform grayscale and coherent texture within the entire region. The image standard deviation decreases from 48 before compensation to 23 after compensation, improving regional consistency.

[0100] Step S124: Extract the image contour edges of the low-brightness connected region map, and set the edge gradient magnitude of the image contour edges to the edge closure threshold to extract the closed road contour and construct the initial road contour map.

[0101] In this embodiment, in the low-brightness connected region map after grayscale filling, an image contour edge extraction operation is performed. First, the image edge is located, and the grayscale change rate of the image in the X and Y directions is calculated to form an edge response map. While extracting the edge contour, the gradient magnitude of each edge pixel is recorded. Then, based on the statistical value of the edge gradient magnitude, a closure threshold of 15 is set. This value is the minimum effective recognition standard for the pixel grayscale change rate. If the gradient values ​​between adjacent edge points are all higher than this threshold, the contour is considered a stable boundary. Contour closure processing is performed in the entire image to construct a closed road contour region. This closed contour is the initial road contour map. Each closed contour in the image forms a structural record with vertex coordinates, bounding box position, and area data during actual processing. This record is subsequently used for fire lane feature identification and spatial projection comparison to form the basic shape structure map for fire identification.

[0102] Optionally, calculating the spatial clearance between obstacles and the road edge in step S2 includes:

[0103] Set the frame difference threshold to 20, perform difference on the time series of consecutive frames of the road image set to obtain the inter-frame difference, and mark the pixels with an inter-frame difference greater than 20 as the foreground region;

[0104] In this embodiment, to determine whether there are dynamic obstacles on the road, pixel-level image difference is performed between every two adjacent image frames and the frame difference threshold is set to 20. That is, if the difference in gray value of the same pixel position in the current frame and the previous frame is greater than 20, the pixel is regarded as a foreground change point and marked as the foreground region. At the same time, the 5% area at the edge of the image is excluded to avoid the influence of errors caused by changes in light or automatic exposure adjustment. The entire foreground extraction operation is limited to the effective road rectangular area in the road image. This area is delineated based on the road boundary line in the initial contour map to shield invalid backgrounds.

[0105] The foreground region mask is calculated based on the inter-frame difference and color difference of the foreground region, thereby identifying dynamic obstacle regions;

[0106] In this embodiment, all foreground changing pixels are aggregated, and the inter-frame brightness difference and RGB color channel difference value are calculated for the formed foreground connected regions as a joint judgment criterion to further eliminate unrealistic dynamic regions caused by lighting, shadows, or reflections. Only when the inter-frame grayscale change is greater than 20 and the sum of the RGB channel change amplitudes is greater than 60 is it considered a valid foreground region. Based on this rule, a foreground mask map is generated and a morphological opening operation (structural element size is 3×3) is performed on it to eliminate isolated noise and form a candidate map of dynamic obstacle regions. This stage ensures that the proposed regions mainly come from real actions such as vehicle movement or express delivery stacking.

[0107] Identify boundary contours in dynamic obstacle regions with dimensions between 300 and 5000 pixels and aspect ratios between 0.5 and 2.5, and record the center coordinates, side length, and orientation angle of the smallest bounding rectangle of the boundary contour to construct an obstacle structure index table;

[0108] In this embodiment, after extracting the dynamic obstacle region, its boundary contour is analyzed. First, regions with fewer than 300 or more than 5000 contour pixels are excluded to remove excessively small noise and excessively large background occlusions. For contours with compliant dimensions, the minimum bounding rectangle is calculated and the aspect ratio of the rectangle is measured. The aspect ratio is required to be between 0.5 and 2.5 to filter obstacle regions with regular shapes, such as ordinary cars, bicycles, temporary tables and benches, and delivery boxes. Contours that meet this condition are further marked, and the center point coordinates, side length (in pixels), and angle information of their bounding rectangles are all structured and stored in the obstacle structure index table. Each record in the table has a corresponding image frame number to support subsequent time matching and dynamic trend analysis.

[0109] The boundary lines on both sides of the road are determined based on the gray-level gradient changes in the upper and lower halves of each image in the road image set.

[0110] In this embodiment, for each frame of image, gray-scale gradient statistics are performed on the upper and lower halves. The upper half is mainly used to detect the left boundary line of the road, and the lower half is mainly used to detect the right boundary line. Pixels with gradient magnitude greater than 25 are set as valid edge points. A top-down sliding window method is used to find the maximum gradient response line of the continuous row of gray-scale abrupt changes as the candidate road boundary line position. In practical applications, the average of 10 consecutive frame boundary lines is used to improve stability. The road boundary line is composed of a set of pixel point sequences. The starting point, ending point and slope value of each line are recorded in the image coordinate system and used for subsequent calculation of the relative distance to obstacles.

[0111] The shortest Euclidean distance from the center coordinates of the obstacle structure index table to the boundary lines on both sides of the road is calculated, and the shortest Euclidean distance is normalized with the overall road width in the traffic road image set to obtain the traffic road gap data.

[0112] In this embodiment, the Euclidean distance from the center point of each obstacle to the left and right road boundary lines is calculated by combining the extracted obstacle structure index table and road boundary line data. The minimum value is then taken as the actual distance from the obstacle to the road boundary. Since the road width varies greatly under different image perspectives, the road width, i.e., the average pixel distance between the two road boundary lines, also needs to be calculated in each frame of the image. This distance value is used as a normalization factor. The obstacle spacing is divided by the current road width to obtain a standardized clearance ratio. This ratio is used to measure the actual impact of obstacles on traffic capacity. When the clearance ratio is less than 0.25, the obstacle is determined to constitute serious obstruction. The above calculation results are summarized to form the final traffic road clearance data, which is used to support subsequent traffic obstruction road image recognition.

[0113] Optionally, identifying dynamic obstacle regions includes:

[0114] Calculate the differences of the foreground region in the R, G, and B channels in adjacent time frames to obtain the RGB channel differences;

[0115] In this embodiment, for the calculation of the RGB channel differences in the foreground region across adjacent time frames, continuous frame color image data is used as input. The values ​​of the red, green, and blue channels for corresponding pixels in each frame are extracted, and the difference matrix of the R, G, and B channels is obtained through pixel-by-pixel subtraction. To ensure the validity of the differences, the difference range is limited to 0 to 255. When the absolute value of the difference between the three RGB channels is less than or equal to 5, it is considered as having no significant change. Pixels whose difference calculations fall outside this range are marked as potential dynamic change areas. In this step, the time interval for difference calculation is strictly between consecutive frames to ensure the timeliness of the time-series data. The frame rate is generally set to 25fps to ensure the smoothness and real-time performance of dynamic detection.

[0116] Extract connected regions from the foreground region where the sum of the RGB channel differences is greater than 50, and construct a color change response map of the connected region.

[0117] In this embodiment, from the RGB channel difference data obtained above, a channel difference sum image is generated by accumulating the three channel differences of each pixel. Connected pixel blocks with a total difference greater than 50 are further selected as candidate regions with significant color changes. The connectivity of connected pixel blocks is determined using 8-connected neighborhood criteria to ensure the continuity and integrity of pixels within the region. For each connected region, a color change response map is automatically constructed, where the pixel values ​​of each connected region in the image represent the intensity of its color change. This color response map structure is represented by a two-dimensional matrix, with the same size as the input image, facilitating rapid location of dynamic regions in subsequent processing steps.

[0118] The foreground region mask is obtained by pixel-by-pixel superimposing of the inter-frame difference and the color change response map;

[0119] In this embodiment, the inter-frame difference image and the color change response map are superimposed pixel by pixel. Specifically, the grayscale values ​​of corresponding pixels in the two images are added together to obtain a comprehensive foreground region mask image. In this mask image, a higher pixel value indicates a more significant dynamic change at that location. To avoid misjudgments due to brightness fluctuations or lighting changes, the pixel values ​​of the mask image are normalized, limiting the value range to between 0 and 1, and a threshold of 0.4 is set as the determination boundary for dynamic regions. Regions with pixel values ​​below this threshold are considered static backgrounds, thereby improving the accuracy of dynamic obstacle recognition.

[0120] Morphological opening is performed on the foreground region mask to remove scale noise interference regions, and connected component extraction is performed on the continuous regions in the foreground region mask after removing scale noise interference regions to obtain the dynamic obstacle region.

[0121] In this embodiment, morphological opening operations are performed on the obtained foreground region mask image to remove small-scale noise. The opening operation uses a circular structuring element with a radius of 3 pixels for morphological processing. First, erosion is performed to remove isolated noise, and then dilation is performed to restore the region shape. After morphological processing, connected component extraction is performed, and the remaining continuous pixel blocks in the mask are analyzed. Isolated regions with an area less than 300 pixels are removed, while larger effective dynamic regions are retained. The final output dynamic obstacle region is stored in the form of a rectangular bounding box, containing boundary coordinates, area, and shape parameters, facilitating subsequent obstacle tracking and classification.

[0122] Figure 3 This is a schematic diagram illustrating the application scenario of step S3 in this invention; for example... Figure 3 As shown, the target area for image analysis within the fire lane area of ​​an urban residential community can be divided into several key areas, including but not limited to: key obstruction identification area 101, typical passage occupancy detection area 102, and traffic bottleneck evolution monitoring area 103.

[0123] Key obstruction identification area 101 may include, but is not limited to, the road space near the main entrance and exit of the fire lane, which is the primary access point for fire trucks to enter and exit the community. Once vehicles occupy the area for a long time, physical barriers are set up, or temporary objects are piled up, it will directly prevent fire trucks from entering the community.

[0124] Typical passageway obstruction detection area 102 may include, but is not limited to, fire lanes near the main entrance or stairwell of residential buildings within the community. This area is often mistakenly used for temporary parking, express delivery, or stacking of goods, and dynamic obstacles may repeatedly appear and disappear.

[0125] The traffic bottleneck evolution monitoring area 103 may include, but is not limited to, the edge passage section at the fire lane exit of the community or a narrow corner. The typical characteristic of this area is that the road width is originally close to the critical passage scale. If any minor road occupation behavior occurs (such as the parking of shared bicycles, the placement of garbage cans, etc.), it may cause the passage of fire trucks to be restricted.

[0126] Optionally, step S3 includes:

[0127] Step S31: Align the fire lane image set with the congested road image set by coordinates, extract the spatially overlapping area, and form a subset of intersecting road images;

[0128] In this embodiment, the image set collected from the fire lane monitoring equipment is aligned in both time and spatial coordinates with the image set of congested roads. Time alignment is achieved using a unified clock and file timestamps, while spatial alignment is based on the camera's installation coordinates (GNSS positioning data) and image edge feature points (such as paving stone seams and manhole cover edges) to construct the projective transformation relationship between the two images. The spatially overlapping areas are extracted as the analysis target, forming a subset of the intersecting road images. Figure 3 Taking area number 101 as an example, this area is located at the entrance and exit of the fire lane, which is a key location with the highest degree of spatial overlap and frequent traffic activities.

[0129] Step S32: Extract the road boundary lines from the subset of intersecting road images, and perform perspective distortion correction on the road boundary lines based on the camera intrinsics and shooting height of the monitoring equipment to obtain the actual intersecting road images;

[0130] In this embodiment, for each image in the subset of intersecting road images, an edge recognition method based on grayscale changes is used to extract the left and right boundary lines of the roads, eliminating interference from shadows and ground markings. Subsequently, based on the intrinsic parameter matrix (focal length f = 4.0 mm, principal point position cx = 960, cy = 540) stored during the deployment of each camera and the installation height h = 2.5 meters, geometric back projection is used to perform perspective correction on the original images, outputting the actual intersecting road images in the affine plane. Figure 3 The road in front of the residential building corresponding to area 102 has curved edges due to lens distortion, and needs to be restored to a standard rectangular passage structure through the above correction process.

[0131] Step S33: Measure the pixel width of each road outline in the actual intersection road image and convert it into the real scene width to obtain the actual width of the intersection roads;

[0132] In this embodiment, multiple vertical measurement line segments are extracted along the transverse direction of the road boundary line in the image after distortion correction (sampled every 0.5 meters). Their pixel spacing is calculated, and the actual width is calculated based on the calibrated spatial transformation ratio (1 pixel ≈ 2 cm). Each measurement data segment is combined with regional location information (number index + relative position coordinates) and recorded in the "Road Transverse Width Table". For example, in... Figure 3 At the corner of the community exit shown in area number 103, the three measured widths are 3.1 meters, 3.0 meters and 2.8 meters respectively, showing a clear trend of narrowing passage.

[0133] Step S34: Extract the standard fire lane width from the fire lane image set, compare it with the actual width of the intersecting roads, calculate the road width difference, and set the width deviation judgment threshold to 0.6 meters. Mark all image areas in the intersecting road image subset where the road width difference is greater than 0.6 meters as suspected obstruction areas.

[0134] In this embodiment, the standard width of the fire lane in the community (e.g., 3.5 meters) is extracted from a preset database and compared with the actual width of each road measured in step S33. The width difference is calculated for each road. For image areas where the width difference is greater than a set threshold (0.6 meters), they are marked as suspected obstruction areas, and spatial labels for suspected areas are established. Figure 3 The minimum width of area 101 was measured to be 2.2 meters, with a width difference of 1.3 meters. It was automatically identified as a "high obstruction risk area" and marked with a "passage interruption" risk level label.

[0135] Step S35: Determine the type of obstruction for suspected obstruction areas to obtain the areas obstructing the passage of fire trucks.

[0136] In this embodiment, for all suspected obstruction areas, the angle along the long side of the bounding rectangle of the area is calculated, and the included angle is calculated with reference to the principal axis direction of the image (i.e., the longitudinal direction of the road). If the angle is less than 15°, the obstruction is determined to be a "straight-ahead obstruction"; if the angle is greater than 45°, it is determined to be a "turning obstruction". Figure 3 Taking area number 102 as an example, this area is close to the main entrance of the residential building, and its outer contour angle is 10°, so it is judged to be a typical "straight-through obstruction"; Figure 3 Area 103 is located at a narrow corner, with its main outline forming an angle of 65° with the main axis of the image. Combined with the fact that the turning radius of the fire truck is 7.5 meters and the distance between the center and the edge of the corner is only 4.8 meters, it does not meet the turning conditions and was ultimately identified as a "turning obstruction".

[0137] Of particular importance is that step S32 specifically includes:

[0138] Edge detection processing is performed on each image in the intersection road image subset to extract high gradient change regions in the images and obtain an initial road boundary candidate edge map;

[0139] In this embodiment, for each image in the subset of intersecting road images, an edge enhancement technique based on grayscale gradient is applied to highlight pixel areas with significant brightness variations. Specifically, the color image is first converted to grayscale, then the gradient magnitude of each pixel is calculated, and a 3×3 Sobel kernel is used to estimate the horizontal and vertical gradients. By setting a gradient threshold of 30, pixels with gradient magnitudes higher than this threshold are selected to form a preliminary edge candidate image. To avoid isolated points due to noise interference, a non-maximum suppression method is further used to refine the edge contours, resulting in continuous and slender edge lines, which initially identify possible road boundaries.

[0140] Line fitting is performed on relatively parallel or linearly distributed edge segments in the initial road boundary candidate edge map to fit the left and right boundary lines of the road in the image, thus obtaining the road boundary line pairs in the image space.

[0141] In this embodiment, straight line shape recognition processing is performed on the edge segments in the initial edge candidate image. This process utilizes a set of edge points within a local neighborhood. Based on the spatial location and orientation information of these points, approximate straight lines are matched using the least squares method to extract relatively parallel and linearly distributed edge segments. The extraction result consists of two main boundary lines, corresponding to the left and right edges of the road, respectively. These two boundary lines are represented in image space as parameters, including the slope and intercept of the line. To improve the robustness of the fitting, the line segment length is limited to no less than 30% of the image width, and the maximum fitting error is limited to no more than 2 pixels, ensuring that the boundary lines accurately cover the road edge area.

[0142] The camera's intrinsic parameters and installation shooting height parameters preset by the monitoring equipment are called to construct a homography transformation matrix from the image coordinate system to the world coordinate system;

[0143] In this embodiment, camera intrinsic parameter data (including focal length, principal point coordinates, and distortion parameters) pre-stored in the monitoring equipment, along with camera height information measured during installation, are used to establish a mapping relationship between the image coordinate system and the actual world coordinate system. Specifically, a 3×3 homography transformation matrix is ​​constructed, which is calculated using multiple control points on the corresponding image and their actual ground coordinates. This transformation ensures that the road boundary line is mapped from the image space with perspective distortion to a distortion-free planar viewpoint, making the measured distances and angles more accurate. The matrix elements during the transformation process are optimized for numerical stability to ensure that the transformation accuracy error is less than 0.05 meters.

[0144] Perspective correction is performed on road boundary line pairs in image space using a homography transformation matrix to obtain a distortion-corrected image;

[0145] The distortion-corrected image is cropped into an image fragment containing only the area enclosed by the road boundary lines, and this image fragment is used as the actual intersection road image.

[0146] In this embodiment, the homography transformation matrix described above is used to perform perspective correction on the road boundary line pairs in the image space, stretching or compressing the distorted areas in the original image to their true proportions. After the transformation, based on the corrected boundary line coordinates, the image is cropped, retaining only the rectangular region enclosed by the road boundary lines as the output. The boundary of this cropped region is strictly determined according to the outermost extension point of the transformed boundary lines to avoid including redundant background information. The size of the cropped image segment is dynamically adjusted, maintaining the aspect ratio consistent with the original road segment to ensure the accuracy and consistency of subsequent analysis and processing. This segment serves as the actual intersection road image, used for further width measurement and obstacle identification.

[0147] Optionally, step S35 includes:

[0148] Step S351: Determine the main direction of the suspected obstruction region in the local polar coordinate system based on the angle pointed to by the long side of the suspected obstruction region, and obtain the main direction of the obstruction region;

[0149] In this embodiment, to determine the main direction of a suspected obstruction area, the length of the longest side of the smallest bounding rectangle of the area's boundary and its corresponding angle value are extracted. Based on a pre-constructed local polar coordinate system (with the origin set at the monitoring equipment installation point), the long side angle is converted into a polar coordinate angle representation. The angle measurement range is set to 0 to 180 degrees, and the angle value increases counterclockwise. The angle between the long side vector and the polar coordinate reference axis is calculated to ensure that the main direction accurately reflects the direction of the obstacle in space. This main direction is used as a key parameter for subsequent obstruction type determination. During operation, the angle measurement error is controlled within ±2 degrees to ensure the accuracy of the judgment.

[0150] Step S352: Use the image principal axis of the intersection road image of the obstruction area to determine the type. If the angle between the main direction of the obstruction area and the image principal axis is less than or equal to 45 degrees, the suspected obstruction area is determined to be a straight obstruction area for fire trucks; if the angle between the main direction of the obstruction area and the image principal axis is greater than 45 degrees, the suspected obstruction area is determined to be a candidate area for turning obstruction.

[0151] In this embodiment, the angle of the main direction of the obstruction area is compared with the principal axis of the actual intersection road image. The principal axis is calculated through the overall road contour direction. Specifically, the set of edge points of the road contour is extracted, and then principal component analysis (PCA) is used to statistically analyze the two-dimensional coordinates of these points to calculate the covariance matrix of the point set. Through eigenvector decomposition of the covariance matrix, the first principal component vector representing the direction of maximum variance of the point set is obtained, and the direction of this vector is defined as the principal axis of the image. This principal axis represents the road extension direction, and the angle range is limited to 0° to 180°. In the judgment logic, if the angle between the two is less than or equal to 45 degrees, it is determined that the direction of the obstruction area is basically consistent with the road travel direction, and therefore it is classified as a straight obstruction area for fire trucks. If the angle exceeds 45 degrees, it is considered to have a large deviation from the road travel direction and is initially classified as a candidate area for turning obstruction. This judgment threshold of 45 degrees is based on the characteristics of fire truck travel trajectory and traffic safety regulations, and is obtained through a large amount of actual scene annotation data to ensure that the judgment result is reasonable and reliable.

[0152] Step S353: Obtain the minimum turning radius of the fire truck through the Internet of Things, and determine the turning accessibility based on the obstacle boundary features in the actual intersection road image corresponding to the candidate turning obstacle area and the minimum turning radius of the fire truck. If the shortest distance from the center of the inner corner of the road to the nearest obstacle boundary in the corresponding actual intersection road image is less than the minimum turning radius of the fire truck, then the candidate turning obstacle area is determined to be the turning obstacle area of ​​the fire truck.

[0153] In this embodiment, the minimum turning radius parameter of the corresponding fire truck is obtained in real time through an IoT interface. This parameter is pre-set based on vehicle type, size, and steering performance and stored in a cloud database. For candidate turning obstruction areas, the coordinates of the corner center are extracted from the actual intersection road image, and the shortest Euclidean distance from the center to the nearest obstacle boundary in that area is measured. If this distance is less than the minimum turning radius of the fire truck, it indicates that the obstacle seriously affects the vehicle's turning passage, and therefore the area is confirmed and marked as a turning obstruction area for the fire truck. During the operation, the error of the shortest distance measurement is controlled within ±0.1 meters to ensure the spatial accuracy of the judgment.

[0154] Step S354: Spatially merge the fire truck straight-through obstruction area and the fire truck turning obstruction area to obtain the fire truck passage obstruction area.

[0155] In this embodiment, the identified straight-ahead obstruction areas and turning obstruction areas for fire trucks are spatially merged. Specifically, the boundary coordinates of the two types of areas are aggregated to generate a unified spatial polygon, eliminating overlaps and gaps to form a continuous passage obstruction area. This area information serves as input data for subsequent obstacle classification, warning, and placement strategies. During the merging operation, spatial topology analysis based on point sets is used to ensure the accuracy and integrity of the merged area's boundaries, with a spatial error during the merging process not exceeding 0.05 meters.

[0156] Optionally, the obstacle classification in step S4 includes:

[0157] Extract the spatial location, bounding rectangle size, angle direction, duration in frames, and motion speed estimate of all obstacles in the area obstructing the passage of fire trucks, and construct a dynamic feature matrix of obstacles;

[0158] In this embodiment, for all obstacles within the area obstructing fire truck passage, their three-dimensional spatial coordinates (obtained using a local polar coordinate system) are collected. Combined with the corresponding image sequence frames, the minimum bounding rectangle parameters of the obstacles are extracted, including the rectangle's width, height, and the angle of its longer side. Through continuous frame image analysis, the duration (frames) of each obstacle's appearance is calculated, and its velocity is estimated based on positional changes, with the velocity unit limited to meters per second. The above data is summarized for each obstacle to form an obstacle dynamic feature matrix. The rows of the matrix represent different obstacles, and the columns represent spatial position (X, Y, Z), dimensions (width, height), orientation angle, time frame length, and velocity estimate, respectively. This matrix structure facilitates subsequent feature fusion and filtering. During operation, the velocity estimation is calculated using the displacement difference between two frames, with the error controlled within ±0.1 meters per second.

[0159] Extract image texture features corresponding to the areas obstructing the passage of fire trucks, identify high contour regularity areas and low contour regularity areas based on image texture features, and filter candidate obstacles for communication devices from the high contour regularity areas by combining the obstacle dynamic feature matrix.

[0160] In this embodiment, image texture features, including texture roughness, uniformity, and edge regularity, are extracted for the image region corresponding to the area obstructing fire truck passage. Specifically, gray-level co-occurrence matrix analysis is used to extract texture contrast and consistency indices. Based on texture regularity, obstacles are divided into high-regularity regions and low-regularity regions. High-regularity regions generally appear as image blocks with fine textures and clear edges, typically corresponding to equipment surfaces or regular objects; low-regularity regions appear as rough textures and irregular edges. Combining the size and motion state information in the obstacle dynamic feature matrix, candidate equipment obstacles with stable shapes and that are stationary or slowly moving are selected from the high-regularity regions for further text feature recognition.

[0161] Identify the image and text features corresponding to candidate obstacles that can communicate with devices, and access the IoT device database for text index matching. If a matching device text index exists, mark the candidate obstacle that can communicate with devices as a communication obstacle.

[0162] In this embodiment, for candidate obstacles containing communicable devices, a text recognition module analyzes the textual identifiers of potential devices in the image. The recognized content includes device number, name, or model information. The recognition result is then queried through an IoT interface to a local or cloud-based device database, which contains device text indexes and corresponding device information. If the recognized text matches the text index of a device record in the database, the obstacle is confirmed as a communicable obstacle, and its status is updated. This process ensures accurate identification of device types and avoids misjudgments. The device database is updated daily to maintain timeliness and accuracy.

[0163] The relative three-dimensional center of the remaining obstacles that are not marked as communicable obstacles in the area obstructing the passage of fire trucks is extracted, and the straight-line distance between the relative three-dimensional center of the remaining obstacles and the main traffic axis in the actual intersection road image is calculated. If the straight-line distance is less than 1.2 meters and the remaining obstacle belongs to a low contour regularity area, the remaining obstacle is determined to be an obstacle to be cleared.

[0164] In this embodiment, the remaining obstacles not marked as communicable obstacles undergo further processing. First, the three-dimensional spatial center coordinates of these obstacles are extracted, and the vertical straight-line distance between them and the defined main traffic axis in the actual intersection road image is calculated. This distance threshold is set at 1.2 meters, meaning that obstacles within this distance are highly likely to obstruct passage. Simultaneously, considering the low-regularity area attribute where the obstacle is located, obstacles that meet the distance threshold and have a rough texture are classified as obstacles to be cleared. This classification facilitates the prioritization of subsequent clearing operations. The distance calculation error in this process is controlled within ±0.05 meters to ensure the accuracy of spatial positioning.

[0165] Optionally, step S5 includes:

[0166] Step S51: Extract the texture continuity and grayscale balance of the ground plane view area in each image of the security area image set, and identify candidate spatial empty areas;

[0167] In this embodiment, within the security area image set, texture continuity and grayscale uniformity are analyzed for the ground plane view area in each frame of the image. Specifically, the image is converted to grayscale, and the ground area is divided into multiple fixed-size sliding window units (e.g., 5×5 pixels). The mean and variance of the grayscale value of each unit are calculated to evaluate its texture continuity and grayscale uniformity indices. A continuity threshold of 0.7 (range 0–1) and a uniformity threshold of 0.6 are set. Only areas that simultaneously meet both indices are marked as candidate spatial free areas. This step eliminates areas with obvious texture breaks or grayscale unevenness, ensuring that the identified free areas truly reflect the continuous and usable space of the ground.

[0168] Step S52: Perform spatial unitization processing on the candidate empty space area, remove spatial units smaller than the set minimum placement area threshold of 1.5 square decimeters, and calculate the centroid coordinates, the side length of the maximum inscribed rectangle and the accessibility of adjacent channels for each remaining empty unit to generate a structured spatial area table.

[0169] In this embodiment, the identified candidate empty spatial regions are processed into spatial units. Based on the image spatial scale, the empty regions are divided into independent spatial units, with unit area measured in square meters. Units with an area less than 1.5 square meters are discarded to avoid fragmented space affecting the placement effect. For the remaining spatial units, their centroid coordinates are calculated as representative values ​​of their spatial location. Simultaneously, based on the boundary point set, the side length of the maximum inscribed rectangle within each unit is calculated as an indicator of the regularity of the spatial shape. Furthermore, the connectivity paths between each spatial unit and adjacent channels are analyzed, and an accessibility score is calculated, ranging from 0 to 1. A higher score indicates easier access to the unit. All parameters are recorded by spatial unit, forming a structured spatial region table for subsequent matching and strategy formulation.

[0170] Step S53: Use the structured spatial region table to perform spatial shape adaptation matching on the communicable obstacles, and add the communication contact number in the device text index to the adaptation matching to obtain the communicable obstacle placement strategy.

[0171] In this embodiment, a structured spatial region table is used to perform shape fit matching on identified communicable obstacles. Shape fit measures the degree of matching between the obstacle's three-dimensional occupancy dimensions and the spatial unit. The ratio of the obstacle's length, width, and height to the side length of the inscribed rectangle of the spatial unit is calculated, with the fit range limited to between 0.7 and 1 to ensure the space can adequately accommodate the obstacle. Furthermore, the communication contact number corresponding to the device's text index in the device database is embedded in the matching results for easier subsequent management and communication. The resulting communicable obstacle placement strategy details the correspondence between obstacles and spatial units, as well as contact methods, achieving precise device placement.

[0172] Step S54: Based on the structured spatial region table, minimize the center-of-gravity distance of the obstacles to be cleared to obtain the obstacle placement strategy;

[0173] In this embodiment, obstacles to be cleared are allocated based on a minimum distance from their center of gravity according to a structured spatial area table. Specifically, the three-dimensional center of gravity coordinates of each obstacle are extracted, and its Euclidean distance to the center of gravity of each spatial unit is calculated. Using the minimum distance principle, obstacles are preferentially allocated to the nearest spatial unit with good accessibility. A maximum allocation distance threshold of 3 meters is set; allocation requests exceeding this distance are marked as abnormal. This allocation strategy optimizes the storage location of obstacles to be cleared, reduces transport distances and passage obstructions, and improves management efficiency.

[0174] Step S55: Spatially merge the communicable obstacle placement strategy with the obstacle placement strategy to be cleared to obtain the obstacle placement strategy, and upload the obstacle placement strategy to the Internet of Things to perform the map visualization task.

[0175] In this embodiment, the strategy for placing communicable obstacles and the strategy for placing obstacles to be cleared are merged to generate a unified obstacle placement strategy. During the merging process, spatial unit allocation conflicts are detected and adjusted to ensure that the spatial allocation is non-overlapping and continuous. The merged strategy information is uploaded to the cloud platform via an IoT communication module, enabling a visual display of the security area map. The map dynamically displays the precise placement location, status, and related contact information of obstacles, assisting managers in on-site command and dispatch, improving emergency response efficiency and on-site order maintenance.

[0176] Of particular importance is that step S51 specifically includes:

[0177] Step S511: Perform perspective correction processing on each frame of the security area image set to restore the approximate top-down view and obtain the orthographic projection image frame sequence.

[0178] In this embodiment, perspective transformation correction is performed on each frame of the original image in the security area image set to restore the ground plane from a top-down perspective. Specifically, the intrinsic parameters and installation height of the monitoring equipment are first extracted, and then a homography transformation matrix is ​​constructed by combining the relative angle information between the camera and the ground. This matrix maps the point coordinates in the image to the real ground coordinate system, realizing the morphological restoration of roads and spatial areas in the image. In actual implementation, the image resolution is moderately adjusted, typically set to 1920×1080 pixels to ensure a balance between the computational efficiency of perspective transformation and detail preservation. The resulting orthographic projection image sequence effectively avoids perspective distortion, providing accurate basic image data for subsequent spatial analysis.

[0179] Step S512: Perform a region sliding window scan on the orthographic projection image frame sequence, and select regions with high continuous texture, smooth boundary changes and grayscale uniformity within the set uniformity threshold range as initial candidate planar regions based on the scan results.

[0180] In this embodiment, a sliding window scan is performed on the orthographically projected image frame sequence to extract planar candidate regions with continuous texture and regular shape. Specifically, the image is divided into fixed-size scanning windows, typically set to 32×32 pixels to ensure detailed capture of region features. For each window region, the standard deviation of grayscale values ​​and edge gradient changes are measured to evaluate the smoothness and continuity of the texture. The preset texture continuity threshold is 0.75, the boundary change smoothness threshold is 0.3, and the grayscale uniformity is controlled between 0.6 and 0.85. Regions meeting these criteria are identified as preliminary candidate planar regions, excluding regions with discontinuous textures or drastic grayscale fluctuations, ensuring that the selected regions possess good ground flatness and visual stability.

[0181] Step S513: Spatial cross-exclusion of the initial candidate planar region and the foreground region mask to obtain the candidate spatial free region;

[0182] In this embodiment, the initially selected candidate planar regions are spatially cross-excluded with the foreground region mask. Specifically, the foreground region mask is a dynamic object mask generated by the previous motion detection module, used to identify areas in the image where moving obstacles exist. The candidate planar regions are screened at the pixel level, and areas overlapping with the foreground mask are removed, ensuring that the finally identified free areas are free from dynamic obstacle interference. This spatial cross-exclusion process is based on image coordinate consistency and achieves rapid determination through mask logic operations. An overlap tolerance threshold of 5% is also set to allow for slight overlap to avoid false rejection and ensure the integrity and accuracy of the spatially free areas.

[0183] Step S514: Perform connected component analysis on the candidate spatial free regions, and connect the isolated small areas in the connected component analysis results with the adjacent regions to generate candidate spatial free regions.

[0184] In this embodiment, connected component analysis is performed on the candidate empty spatial regions after cross-elimination to integrate fragmented regions and improve the coherence of the empty regions. By identifying groups of connected pixels, isolated small areas (with an area threshold set to less than 0.3 square meters) are assessed for their spatial proximity to adjacent large areas. If the adjacent distance is less than 10 pixels, they are merged into the nearest main region. This operation effectively avoids spatial breaks caused by occlusion or changes in lighting, enhancing the complete structure of the empty regions. The resulting candidate empty spatial regions have higher spatial connectivity and practical value, providing a reliable spatial basis for subsequent obstacle placement.

[0185] Optionally, this specification also provides an IoT-based digital visualization management system for security equipment, used to execute the IoT-based digital visualization management method for security equipment as described above. The IoT-based digital visualization management system for security equipment includes:

[0186] The image segmentation module is used to acquire a set of security area images, segment the security area image set into road image regions, and obtain a set of fire lane images and a set of traffic road images;

[0187] The obstruction road recognition module is used to calculate the spatial gap between obstacles and road edges on the road based on the road image set, and obtain road gap data; and to identify the obstruction road image set in the road image set based on the road gap data.

[0188] The fire truck access obstruction area identification module is used to compare the road width difference between the intersection road between the fire lane image set and the access obstruction road image set, and identify the fire truck access obstruction area based on the road width difference. The fire truck access obstruction area includes the fire truck turning obstruction area and the fire truck straight-through obstruction area.

[0189] The obstacle classification module is used to identify the characteristics of obstacles in the area where fire trucks are obstructing passage using a set of images of the passing roads, and classify the obstacles into communicable obstacles and obstacles to be cleared based on the characteristics of the obstacles.

[0190] The obstacle placement module is used to identify vacant areas based on the security area image set, allocate placement areas of vacant areas for communicable obstacles and obstacles to be cleared, obtain obstacle placement strategies, and upload the obstacle placement strategies to the Internet of Things to perform map visualization tasks.

[0191] Therefore, the embodiments should be considered as exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of the equivalents of the application are intended to be included within the invention.

[0192] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features of the invention herein.

Claims

1. A digital visualization management method for security equipment based on the Internet of Things, characterized in that, Includes the following steps: Step S1: Collect a set of security area images, segment the security area image set into road image regions, and obtain a set of fire lane images and a set of traffic road images; Step S2: Calculate the spatial gap between obstacles and road edges on the road based on the road image set to obtain road gap data; identify the road obstruction image set in the road image set based on the road gap data; Step S3: Compare the road width difference between the intersection of the fire lane image set and the road obstruction image set, and identify the fire truck passage obstruction area based on the road width difference. The fire truck passage obstruction area includes the fire truck turning obstruction area and the fire truck straight passage obstruction area. Step S4: Use the image set of the passageway to identify the characteristics of obstacles in the area obstructing the passage of fire trucks, and classify the obstacles into communicable obstacles and obstacles to be cleared according to the characteristics of the obstacles; Step S5: Identify vacant areas based on the security area image set, allocate placement areas for communicable obstacles and obstacles to be cleared, obtain an obstacle placement strategy, and upload the obstacle placement strategy to the Internet of Things to perform map visualization tasks.

2. The method for digital visualization management of security equipment based on the Internet of Things according to claim 1, characterized in that, Step S1 includes: Step S11: Collect image sets of the security area through the pre-deployed monitoring equipment in the security area; Step S12: Extract low-brightness connected regions from the security area image and merge small occluded areas within the low-brightness connected regions to obtain the initial road outline map; Step S13: Identify areas in the initial road outline map that contain red lines, yellow diagonal grid lines, and fire lane text signs as candidate fire lane image segments, and use the remaining areas in the initial road outline map as traffic road image segments; Step S14: Construct a local polar coordinate system with the installation point of the monitoring equipment as the origin; Step S15: Using the local polar coordinate system, the angles of the candidate fire lane image segments and the traffic road image segments are calibrated and their positions are recorded to obtain the fire lane image set and the traffic road image set.

3. The method for digital visualization management of security equipment based on the Internet of Things according to claim 2, characterized in that, Step S12 includes: Step S121: Extract the brightness channel of each frame image in the security area image set, set the brightness threshold to 90, and extract the pixels with brightness below 90 in each frame image to form a low brightness pixel mask image. Step S122: Identify continuous low-brightness regions in the low-brightness pixel mask image and filter out isolated regions with an area of ​​less than 200 pixels to obtain candidate road segments. Step S123: Set the maximum interpolation radius to 10 pixels, perform pixel value estimation on the areas with obvious gray-level changes in the candidate road segments, and use the pixel value estimation results to fill in the areas with obvious gray-level changes to obtain a low-brightness connected region map; Step S124: Extract the image contour edges of the low-brightness connected region map, and set the edge gradient magnitude of the image contour edges to the edge closure threshold to extract the closed road contour and construct the initial road contour map.

4. The method for digital visualization management of security equipment based on the Internet of Things according to claim 1, characterized in that, Step S2, calculating the spatial clearance between obstacles and the road edge on the roadway, includes: Set the frame difference threshold to 20, perform difference on the time series of consecutive frames of the road image set to obtain the inter-frame difference, and mark the pixels with an inter-frame difference greater than 20 as the foreground region; The foreground region mask is calculated based on the inter-frame difference and color difference of the foreground region, thereby identifying dynamic obstacle regions; Identify boundary contours in dynamic obstacle regions with dimensions between 300 and 5000 pixels and aspect ratios between 0.5 and 2.5, and record the center coordinates, side length, and orientation angle of the smallest bounding rectangle of the boundary contour to construct an obstacle structure index table; The boundary lines on both sides of the road are determined based on the gray-level gradient changes in the upper and lower halves of each image in the road image set. The shortest Euclidean distance from the center coordinates of the obstacle structure index table to the boundary lines on both sides of the road is calculated, and the shortest Euclidean distance is normalized with the overall road width in the traffic road image set to obtain the traffic road gap data.

5. The method for digital visualization management of security equipment based on the Internet of Things according to claim 4, characterized in that, The areas for identifying dynamic obstacles include: Calculate the differences of the foreground region in the R, G, and B channels in adjacent time frames to obtain the RGB channel differences; Extract connected regions from the foreground region where the sum of the RGB channel differences is greater than 50, and construct a color change response map of the connected region. The foreground region mask is obtained by pixel-by-pixel superimposing of the inter-frame difference and the color change response map; Morphological opening is performed on the foreground region mask to remove scale noise interference regions, and connected component extraction is performed on the continuous regions in the foreground region mask after removing scale noise interference regions to obtain the dynamic obstacle region.

6. The method for digital visualization management of security equipment based on the Internet of Things according to claim 1, characterized in that, Step S3 includes: Step S31: Align the fire lane image set with the congested road image set by coordinates, extract the spatially overlapping area, and form a subset of intersecting road images; Step S32: Extract the road boundary lines from the subset of intersecting road images, and perform perspective distortion correction on the road boundary lines based on the camera intrinsics and shooting height of the monitoring equipment to obtain the actual intersecting road images; Step S33: Measure the pixel width of each road outline in the actual intersection road image and convert it into the real scene width to obtain the actual width of the intersection roads; Step S34: Extract the standard fire lane width from the fire lane image set, compare it with the actual width of the intersecting roads, calculate the road width difference, and set the width deviation judgment threshold to 0.6 meters. Mark all image areas in the intersecting road image subset where the road width difference is greater than 0.6 meters as suspected obstruction areas. Step S35: Determine the type of obstruction for suspected obstruction areas to obtain the areas obstructing the passage of fire trucks.

7. The method for digital visualization management of security equipment based on the Internet of Things according to claim 6, characterized in that, Step S35 includes: Step S351: Determine the main direction of the suspected obstruction region in the local polar coordinate system based on the angle pointed to by the long side of the suspected obstruction region, and obtain the main direction of the obstruction region; Step S352: Use the image principal axis of the intersection road image of the obstruction area to determine the type. If the angle between the main direction of the obstruction area and the image principal axis is less than or equal to 45 degrees, the suspected obstruction area is determined to be a straight obstruction area for fire trucks; if the angle between the main direction of the obstruction area and the image principal axis is greater than 45 degrees, the suspected obstruction area is determined to be a candidate area for turning obstruction. Step S353: Obtain the minimum turning radius of the fire truck through the Internet of Things, and determine the turning accessibility based on the obstacle boundary features in the actual intersection road image corresponding to the candidate turning obstacle area and the minimum turning radius of the fire truck. If the shortest distance from the center of the inner corner of the road to the nearest obstacle boundary in the corresponding actual intersection road image is less than the minimum turning radius of the fire truck, then the candidate turning obstacle area is determined to be the turning obstacle area of ​​the fire truck. Step S354: Spatially merge the fire truck straight-through obstruction area and the fire truck turning obstruction area to obtain the fire truck passage obstruction area.

8. The method for digital visualization management of security equipment based on the Internet of Things according to claim 1, characterized in that, Step S4 involves classifying obstacles into different types, including: Extract the spatial location, bounding rectangle size, angle direction, duration in frames, and motion speed estimate of all obstacles in the area obstructing the passage of fire trucks, and construct a dynamic feature matrix of obstacles; Extract image texture features corresponding to the areas obstructing the passage of fire trucks, identify high contour regularity areas and low contour regularity areas based on image texture features, and filter candidate obstacles for communication devices from the high contour regularity areas by combining the obstacle dynamic feature matrix. Identify the image and text features corresponding to candidate obstacles that can communicate with devices, and access the IoT device database for text index matching. If a matching device text index exists, mark the candidate obstacle that can communicate with devices as a communication obstacle. The relative three-dimensional center of the remaining obstacles that are not marked as communicable obstacles in the area obstructing the passage of fire trucks is extracted, and the straight-line distance between the relative three-dimensional center of the remaining obstacles and the main traffic axis in the actual intersection road image is calculated. If the straight-line distance is less than 1.2 meters and the remaining obstacle belongs to a low contour regularity area, the remaining obstacle is determined to be an obstacle to be cleared.

9. The method for digital visualization management of security equipment based on the Internet of Things according to claim 1, characterized in that, Step S5 includes: Step S51: Extract the texture continuity and grayscale balance of the ground plane view area in each image of the security area image set, and identify candidate empty spatial areas; Step S52: Perform spatial unitization processing on the candidate empty space area, remove spatial units smaller than the set minimum placement area threshold of 1.5 square decimeters, and calculate the centroid coordinates, the side length of the maximum inscribed rectangle and the accessibility of adjacent channels for each remaining empty unit to generate a structured spatial area table. Step S53: Use the structured spatial region table to perform spatial shape adaptation matching on the communicable obstacles, and add the communication contact number in the device text index to the adaptation matching to obtain the communicable obstacle placement strategy. Step S54: Based on the structured spatial region table, minimize the center-of-gravity distance of the obstacles to be cleared to obtain the obstacle placement strategy; Step S55: Spatially merge the communicable obstacle placement strategy with the obstacle placement strategy to be cleared to obtain the obstacle placement strategy, and upload the obstacle placement strategy to the Internet of Things to perform the map visualization task.

10. A digital visualization management system for security equipment based on the Internet of Things, characterized in that, For executing the IoT-based digital visualization management method for security equipment as described in claim 1, the IoT-based digital visualization management system for security equipment includes: The image segmentation module is used to acquire a set of security area images, segment the security area image set into road image regions, and obtain a set of fire lane images and a set of traffic road images; The obstruction road recognition module is used to calculate the spatial gap between obstacles and road edges on the road based on the road image set, and obtain road gap data; and to identify the obstruction road image set in the road image set based on the road gap data. The fire truck access obstruction area identification module is used to compare the road width difference between the intersection road between the fire lane image set and the access obstruction road image set, and identify the fire truck access obstruction area based on the road width difference. The fire truck access obstruction area includes the fire truck turning obstruction area and the fire truck straight-through obstruction area. The obstacle classification module is used to identify the characteristics of obstacles in the area where fire trucks are obstructing passage using a set of images of the passing roads, and classify the obstacles into communicable obstacles and obstacles to be cleared based on the characteristics of the obstacles. The obstacle placement module is used to identify vacant areas based on the security area image set, allocate placement areas of vacant areas for communicable obstacles and obstacles to be cleared, obtain obstacle placement strategies, and upload the obstacle placement strategies to the Internet of Things to perform map visualization tasks.

Images