First-aid resource information early warning method and system based on AI identification

Abstract

The invention belongs to the technical field of image processing, and particularly relates to a first-aid resource information early warning method and system based on AI identification, and the method comprises the steps: building a monitoring scene coordinate system, calibrating an operation key point of a first-aid facility, and setting a region of interest; collecting a video stream in real time, extracting a foreground target, and performing attribute preliminary screening according to the morphological regularity to determine a normalized adjustment coefficient; constructing a spatial hindering potential energy model containing distance potential energy based on the distance from the mass center of the foreground target to the operation key point, and calculating an instantaneous spatial hindering index in combination with a normalized adjustment coefficient; and calculating the risk reliability according to the fluctuation change rate of the instantaneous spatial obstruction index by using dynamic and static modal decoupling logic, and triggering early warning when the risk reliability exceeds a threshold value. According to the invention, by introducing the logarithmic potential energy field and the stability weight factor, accurate assessment of distance harm of the shelter and effective filtering of dynamic interference are realized, and the accuracy of first-aid resource shielding early warning is significantly improved.

Description

Technical Field

[0001] This invention relates to the field of image processing technology. More specifically, this invention relates to an AI-based method and system for early warning of emergency medical resource information. Background Technology

[0002] With the development of urban public safety systems, emergency medical facilities such as AED (Automated External Defibrillator) cabinets, mini fire stations, and wall-mounted first aid kits are being deployed more and more widely in public places. To ensure that these facilities can be easily accessed in emergencies and to prevent them from being obstructed by goods, vehicles, or other debris, real-time monitoring of their surrounding environment is necessary.

[0003] The current mainstream emergency facility monitoring solutions mainly rely on background subtraction in fixed areas or general target detection algorithms such as the YOLO series. The core logic is that when a foreign object appears in the defined monitoring area (ROI) and the percentage of pixel change exceeds the set threshold and continues for a certain period of time, the system determines that the passage is blocked and triggers an alarm.

[0004] However, existing technologies have significant limitations in complex urban public environments. Due to the lack of consideration for spatial potential energy, simply calculating the obstruction area cannot identify the specific impact of an object's location on the accessibility of emergency facilities. For example, a small sticker close to a handle may pose a greater hazard than a large box far from the cabinet, leading to a serious risk of missed detections. Secondly, existing technologies are prone to confusing static and dynamic modes. In crowded places, the natural shaking caused by pedestrians leaning against, making phone calls, or cleaners wiping equipment can cause continuous pixel changes, which can easily be misjudged by the system as object stacking, resulting in a large number of invalid work orders. In addition, existing technologies lack morphological logic and have difficulty distinguishing between regular objects such as express delivery boxes and irregular human bodies, resulting in a lack of targeted monitoring. Summary of the Invention

[0005] To address the technical problems of existing technologies being unable to effectively distinguish the degree of harm caused by the location of an object obstructing access and the high false alarm rate caused by easily confusing dynamic human movement with static object stacking, this invention provides solutions in the following aspects.

[0006] In a first aspect, the present invention provides an AI-based early warning method for emergency medical resource information, comprising: establishing a monitoring scene coordinate system; calibrating the center point of the operating handle of the emergency medical facility in the image as a potential energy field singularity; dividing the physical contour region of the emergency cabinet as a region of interest; and simultaneously collecting the average ambient brightness value as an ambient lighting reference; real-time access to the video stream; using a Gaussian mixture background modeling algorithm to remove the background to extract the foreground moving target; performing morphological processing on the foreground moving target to form a complete connected component; calculating the real area ratio of the connected component; performing initial attribute screening on the foreground moving target based on the real area ratio to determine a normalization adjustment coefficient; and calculating the geometric... The centroid is determined by calculating the instantaneous spatial obstacle index generated by the object in the current frame using a preset logarithmic potential energy field model, based on the straight-line distance from the geometric centroid to the center point of the operating handle, the normalized adjustment coefficient, and the pixel area of ​​the connected region. The instantaneous spatial obstacle index is positively correlated with the normalized adjustment coefficient and the pixel area, and negatively correlated with the straight-line distance. The fluctuation rate of change is calculated based on the instantaneous spatial obstacle index of the current frame and the previous frame. A stability weight factor is constructed using the fluctuation rate of change. The risk confidence at the current moment is iteratively calculated by combining the risk confidence accumulated at the previous moment and the natural decay constant. When the risk confidence exceeds a preset threshold, a graded warning is executed.

[0007] This invention constructs a logarithmic model that includes distance potential energy, which can assign different obstacle weights according to the distance between the object and the key operation point, realizing the spatial potential energy assessment that the closer the distance, the greater the harm, effectively solving the defect that the traditional area threshold method cannot assess the location hazard.

[0008] Preferably, the step of performing initial attribute screening on the foreground moving target based on the actual area ratio to determine the normalization adjustment coefficient includes: calculating the actual area ratio of the connected component, where the actual area ratio is the ratio of the area of ​​the connected component to the area of ​​its convex hull; determining whether the actual area ratio is lower than a preset actual area ratio threshold; if it is lower than the preset actual area ratio threshold, it is marked as a low-risk suspected target, and the value of the normalization adjustment coefficient is lowered; if it is not lower than the preset actual area ratio threshold, it is marked as a high-risk suspected target, and the value of the normalization adjustment coefficient is maintained or increased.

[0009] Preferably, the step of calculating the instantaneous spatial obstacle index generated by an object in the current frame using a preset logarithmic potential energy field model includes: calculating the ratio of the potential energy radiation radius to the distance factor, and taking the natural logarithm of the ratio by adding 1 to obtain the distance attenuation factor; wherein the distance factor is composed of the sum of the pixel distance from the object's centroid to the emergency cabinet handle and the safety buffer constant; and multiplying the normalization adjustment coefficient, the pixel area of ​​the foreground connected region, and the distance attenuation factor to obtain the instantaneous spatial obstacle index.

[0010] The natural logarithm function is used to characterize the attenuation of physical field strength, so that the resistance exponent rises smoothly when the object is close to the handle and approaches zero when it is far away. A safety buffer constant is used to prevent calculation errors where the denominator is zero.

[0011] Preferably, the normalization adjustment coefficient is dynamically adjusted based on the total image resolution and the morphological regularity of the foreground moving target. The morphological regularity is characterized by the real volume ratio and is used to suppress the weight of irregular objects.

[0012] By using morphological regularity to initially screen targets and dynamically adjust the coefficients, prior knowledge can be introduced at an early stage. Irregular targets such as suspected human bodies will have their calculation weight automatically reduced, further reducing the false alarm rate caused by short-term personnel stays or operations from the source.

[0013] Preferably, the step of calculating the fluctuation change rate based on the instantaneous spatial obstruction index of the current frame and the previous frame includes: reading the instantaneous spatial obstruction index calculated in the previous frame and the instantaneous spatial obstruction index of the current frame; and calculating the absolute value of the difference between the instantaneous spatial obstruction index of the current frame and the instantaneous spatial obstruction index of the previous frame as the fluctuation change rate.

[0014] Preferably, the iterative calculation of the risk confidence at the current moment includes: calculating a stability weighting factor using an exponential function, wherein the exponential part of the stability weighting factor is the negative value of the product of the stability sensitivity coefficient and the volatility change rate; multiplying the instantaneous spatial obstacle index of the current frame by the stability weighting factor to obtain the risk increment of the current frame; adding the risk confidence accumulated at the previous moment to the risk increment of the current frame, and subtracting the natural decay constant to obtain the final risk confidence accumulated at the current moment.

[0015] By introducing an exponential stability weighting factor, the system achieves accurate early warning by making the weighting factor close to zero for dynamic targets such as pedestrians, whose volatility is relatively high, thus preventing the index from accumulating; while for stationary objects, whose volatility is extremely low, the weighting factor close to 1, allowing the risk confidence to accumulate rapidly.

[0016] Preferably, when calculating the risk confidence level, a boundary condition limiting step is further included: if the calculated risk confidence level is less than zero, the risk confidence level is forcibly assigned to zero; if the calculated risk confidence level is greater than the set saturation upper limit, the risk confidence level is forcibly assigned to the saturation upper limit.

[0017] Preferably, the execution of the graded early warning includes: when the risk confidence level is greater than a first preset ratio threshold and continues to rise, triggering a level one early warning and playing a voice prompt; when the risk confidence level is greater than a second preset ratio threshold, determining it as substantial occlusion, triggering a level two alarm, capturing the current frame and overlaying a heat map of the risk confidence level and the instantaneous spatial obstacle index, and pushing it to the management cloud platform.

[0018] A tiered early warning mechanism is adopted. For minor or initial risks, only local voice prompts are provided to effectively dissuade unintentional obstruction behavior. For serious risks, remote alarms are triggered and visual heat map evidence is pushed, which facilitates rapid assessment and evidence collection by managers, thereby improving the efficiency and pertinence of emergency management.

[0019] Preferably, the straight-line distance is Euclidean distance, and the center point of the operating handle is the singularity of the potential energy field, which is the area with the highest protection level.

[0020] Secondly, the present invention provides an emergency medical resource information early warning system based on AI recognition, including a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, the aforementioned emergency medical resource information early warning method based on AI recognition is implemented.

[0021] By adopting the above technical solution, a computer program is generated from the aforementioned AI-based emergency resource information early warning method and stored in a memory so that it can be loaded and executed by a processor. This allows for the creation of a terminal device based on the memory and processor, making it convenient to use.

[0022] This invention constructs a logarithmic potential energy field model based on the distance between the geometric centroid and the key operation point, mapping the physical position of the obstructing object to a nonlinear spatial obstacle index. It automatically assigns differentiated weights based on the distance of the object from the operation part of the emergency facility, accurately assessing the actual degree of obstruction to facility access at different locations. This solves the problem that the traditional area threshold method cannot assess the locational hazard, and effectively achieves key protection of critical operation areas.

[0023] Furthermore, by monitoring the fluctuation rate of the instantaneous obstruction index, the system automatically suppresses the risk accumulation of human targets with minute movement characteristics and quickly responds to the occlusion behavior of stationary objects, effectively eliminating false alarms caused by short-term personnel stays. While ensuring sensitive detection of illegal stacking behavior, it significantly improves the robustness of the early warning system. Attached Figure Description

[0024] The above and other objects, features, and advantages of exemplary embodiments of the present invention will become readily apparent upon reading the following detailed description with reference to the accompanying drawings. In the drawings, several embodiments of the invention are illustrated by way of example and not limitation, and like or corresponding reference numerals denote like or corresponding parts, wherein: Figure 1 This is a flowchart illustrating an AI-based emergency resource information early warning method according to the present invention; Figure 2 This is a schematic diagram of the risk accumulation response mechanism in an embodiment of the present invention; Figure 3 This is a schematic diagram illustrating the potential energy field distribution model that hinders the operation of emergency medical resources. Detailed Implementation

[0025] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0027] This invention discloses an AI-based early warning method for emergency medical resource information, referring to... Figure 1 This includes steps S1-S4: S1. Establish a coordinate system for the monitoring scene, mark the center point of the operating handle of the emergency facility in the image as the singularity of the potential energy field, and divide the physical contour area of ​​the emergency cabinet as the region of interest. At the same time, collect the average ambient brightness value as the ambient lighting reference.

[0028] In an optional embodiment, a basic image of the emergency medical facility can be acquired via a front-end camera, and a two-dimensional pixel coordinate system can be constructed in the image processing unit. Specifically, the center point of the operating handle of the emergency medical facility is marked in the image; this point is a singularity of the potential energy field, i.e., the area with the highest protection level. Simultaneously, the physical outline area of ​​the emergency medical cabinet is set as the Region of Interest (ROI), and the average brightness value of the current environment is collected as an ambient lighting reference for subsequent removal of false triggers caused by sudden changes in lighting.

[0029] For example, in a 1920x1080 resolution monitoring screen, the first aid cabinet is located in the center of the screen. After calibration, the coordinates of the center point of the operating handle are (960, 540), and the ROI area is set as a rectangle covering the entire front of the first aid cabinet.

[0030] In this way, by accurately calibrating the key point of the operating handle, the foundation is laid for the subsequent construction of a spatial potential energy field centered on that point, enabling the system to perceive the differences in importance of different locations.

[0031] S2. Real-time access to the video stream, use Gaussian mixture background modeling algorithm to remove the background to extract the foreground moving target, perform morphological processing on the foreground moving target to form a complete connected component, calculate the real area ratio of the connected component, and perform initial attribute screening on the foreground moving target based on the real area ratio to determine the normalization adjustment coefficient.

[0032] In an optional embodiment, a video stream is accessed in real time, and each frame of the image is processed. The background is removed using a Gaussian Mixture Background Modeling (GMM) algorithm to extract the foreground moving target. Dilation and closure operations are then performed on the foreground target to fill the holes inside the target and the broken edges of the connection, forming a complete connected domain.

[0033] Further, an initial screening based on morphological regularity is performed, specifically by calculating the actual volume ratio of connected components, i.e., the ratio of the area of ​​a connected component to the area of ​​its convex hull. The logic is that due to their irregular shapes, human bodies typically have a low actual volume ratio, while illegally piled objects such as boxes and vehicles usually have a high actual volume ratio. By setting a preset actual volume ratio threshold, for example, 0.5, targets with an actual volume ratio below this threshold are marked as low-risk suspected targets, and their subsequent weight coefficients (normalization adjustment coefficients) are reduced; while targets with an actual volume ratio above this threshold are marked as high-risk suspected targets, and their corresponding normalization adjustment coefficients are maintained or increased.

[0034] For example, in a certain detection, a foreground connected component with a pixel area of ​​5000 and a calculated convex hull area of ​​8000 are extracted. The actual area ratio is 5000 / 8000 = 0.625. Since 0.625 > 0.5, the system marks it as a high-risk suspected target, speculating that it may be a regular object such as a box, and sets a higher normalization adjustment coefficient. If another connected component with an actual area ratio of 0.4 is extracted, it is speculated that it may be a human body being leaned on, and it is marked as a low-risk suspected target, and the corresponding normalization adjustment coefficient is reduced.

[0035] Thus, by classifying targets based on morphological features and determining the normalization adjustment coefficient accordingly, prior knowledge can be introduced in the early stages, reducing the possibility of false alarms from non-rigid objects such as human bodies in subsequent calculations.

[0036] S3. Calculate the geometric centroid of the connected component. Based on the straight-line distance from the geometric centroid to the center point of the operation handle, the normalization adjustment coefficient, and the pixel area of ​​the connected component, calculate the instantaneous spatial obstacle index generated by the object in the current frame using the preset logarithmic potential energy field model. The instantaneous spatial obstacle index is positively correlated with the normalization adjustment coefficient and the pixel area, and negatively correlated with the straight-line distance.

[0037] In an optional embodiment, the total number of pixels in the foreground connected region can be obtained, and the geometric centroid of the connected region can be calculated. The straight-line distance from the centroid to the center point of the operation handle can be calculated using the Euclidean distance formula. Combined with the normalization adjustment coefficient and the pixel area of ​​the connected region, the instantaneous spatial hindrance index generated by the object in the current frame can be calculated using a preset logarithmic potential energy field model. The instantaneous spatial hindrance index satisfies the following relationship: in, The instantaneous spatial hindrance index generated by objects in the current frame is calculated. The pixel area of ​​the foreground connected region; The pixel distance from the object's center of mass to the first aid cabinet handle; This is a safety buffer constant, for example, with a value of 20. Its function is to prevent the denominator from becoming zero when the object is close to the handle, which would cause the program to crash. At the same time, it adjusts the sensitivity saturation curve of the potential energy field at extremely close distances. The radius of potential energy radiation is, for example, 1 / 3 of the diagonal length of the image, which defines the effective range of influence of the potential energy field. The natural logarithm function is used to characterize the decay characteristics of physical field strength. When the value decreases, the fractional term increases, resulting in a smooth increase in the logarithmic value. Adding 1 ensures that when the object is at infinity... That is, the instantaneous spatial resistance index is 0; The normalization adjustment coefficient is dynamically adjusted based on the total image resolution and morphological regularity.

[0038] For example, suppose , object area , When an object is pressed against the handle, the distance between the object's center of mass and the handle is... pixels, then When an object moves away from the handle, the distance between the object's center of mass and the handle... Pixels As can be seen, although the objects have the same area, due to the difference in distance, the instantaneous spatial resistance index generated by the object close to the handle is much higher than that of the object far from the handle.

[0039] Thus, by constructing a logarithmic potential field model, the degree of obstruction can be nonlinearly assessed based on the distance between the object and the key parts, ensuring the key areas are protected and meeting the actual needs of emergency medical facilities.

[0040] S4. Calculate the volatility change rate based on the instantaneous spatial obstacle index of the current frame and the previous frame, construct a stability weight factor using the volatility change rate, and iteratively calculate the risk confidence at the current moment by combining the risk confidence accumulated at the previous moment and the natural decay constant. When the risk confidence exceeds the preset threshold, execute a graded warning.

[0041] In an optional embodiment, a buffer queue can be set up to read the instantaneous spatial obstruction index calculated in the previous frame and the instantaneous spatial obstruction index of the current frame, and the absolute value of the difference between the instantaneous spatial obstruction index of the current frame and the instantaneous spatial obstruction index of the previous frame can be used as the fluctuation rate to reflect the stability of the obstruction.

[0042] Furthermore, a stability weighting factor can be constructed using the volatility rate of change. This, combined with the accumulated risk confidence from the previous time step and the natural decay constant, is used to iteratively calculate the risk confidence at the current time step. The iterative cumulative relationship for risk confidence is as follows: in, For the current moment The accumulated final risk confidence level is the sole basis for the system to determine whether to issue an alarm. The risk confidence accumulated in the previous moment; This is a stability weighting factor, with a value between 0 and 1. This is the stability sensitivity coefficient, and for example, it is set to 5. The calculated rate of change of volatility; This is the natural decay constant, exemplified by a value of 0.5. If the obstruction is removed, the system needs to... Allow the accumulated risk value to automatically drop back to 0.

[0043] In this optional embodiment, in order to prevent numerical anomalies from occurring during long-term operation of the formula, the following boundary condition limiting steps can be performed: if the calculated risk confidence is less than zero, the risk confidence is forcibly assigned to zero; if the calculated risk confidence is greater than the set saturation upper limit, for example, if the saturation upper limit is 100, the risk confidence is forcibly assigned to the saturation upper limit.

[0044] For example, suppose It is 30, the previous moment. If it is a static stack of objects such as boxes, the objects are stationary. Extremely stable, assuming Then the stability weighting factor is , As can be seen, the risk confidence level rapidly increased from 10 to 38, which the system identified as a high-risk accumulation. If the interference is dynamic, such as a pedestrian, the person may be in front of the counter, but their body may be swaying. The fluctuations are large, assuming Then the stability weighting factor is , It can be seen that although The risk is very high, but due to its large volatility and extremely low weight, the risk confidence level actually decreases.

[0045] Furthermore, this can be achieved through real-time monitoring. The system uses numerical values ​​to generate tiered warnings. When the risk confidence level exceeds the first preset threshold (e.g., 30) and continues to rise, a Level 1 warning is triggered, and a voice prompt, "Please do not obstruct emergency facilities," is played. When the risk confidence level exceeds the second preset threshold (e.g., 80), it is determined to be a substantial obstruction, triggering a Level 2 alarm. The system can immediately capture the current frame and overlay a heat map of the risk confidence level and instantaneous spatial obstruction index, which is then pushed to the management cloud platform via a 4G / 5G network.

[0046] like Figure 2 As shown in the figure, it is a schematic diagram of the risk accumulation response mechanism in the embodiment of the present invention. It can be seen that whether it is a pedestrian or an object, the original signal captured by the sensor remains at a high level and is difficult to distinguish. However, the risk accumulation confidence curve of the present invention is suppressed to a low level in the dynamic interference scenario, and rises rapidly to break through the threshold in the static occlusion scenario, which reflects the technical effect of the present invention in filtering dynamic false alarms and locking static occlusion.

[0047] like Figure 3 The diagram shown is a schematic representation of the potential energy field distribution model that hinders the operation of emergency resources. It can be seen that the thermal color gradually lightens from the center point outwards. Once an object enters the center point area, the spatial resistance potential energy intensity will rise rapidly, demonstrating that the system has the spatial perception capability to protect key parts.

[0048] In this way, it can effectively distinguish between a slightly swaying human body and a completely stationary object, eliminating false alarms caused by common scenarios such as pedestrians leaning against objects, while ensuring sensitive detection of real illegal stacking behavior.

[0049] This invention also discloses an AI-based emergency medical resource information early warning system, including a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, an AI-based emergency medical resource information early warning method according to this invention is implemented.

[0050] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

[0051] In the description of this specification, "multiple" or "several" means at least two, such as two, three or more, unless otherwise expressly and specifically defined.

[0052] While this specification has shown and described numerous embodiments of the invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Many modifications, alterations, and alternatives will occur to those skilled in the art without departing from the spirit and essence of the invention. It should be understood that various alternatives to the embodiments of the invention described herein may be employed in the practice of this invention.

Claims

1. A method for early warning of emergency medical resource information based on AI recognition, characterized in that, include: Establish a coordinate system for the monitoring scene, mark the center point of the operating handle of the emergency facility in the image as the singularity of the potential energy field, and divide the physical contour area of ​​the emergency cabinet as the region of interest. At the same time, collect the average ambient brightness value as the ambient lighting reference. The video stream is accessed in real time. The background is removed using a Gaussian mixture background modeling algorithm to extract the foreground moving target. The foreground moving target is morphologically processed to form a complete connected component. The real area ratio of the connected component is calculated. The foreground moving target is initially screened for attributes based on the real area ratio to determine the normalization adjustment coefficient. Calculate the geometric centroid of the connected region. Based on the straight-line distance from the geometric centroid to the center point of the operation handle, the normalization adjustment coefficient, and the pixel area of ​​the connected region, calculate the instantaneous spatial obstacle index generated by the object in the current frame using a preset logarithmic potential energy field model. The instantaneous spatial obstacle index is positively correlated with the normalization adjustment coefficient and the pixel area, and negatively correlated with the straight-line distance. The volatility change rate is calculated based on the instantaneous spatial obstacle index of the current frame and the previous frame. A stability weight factor is constructed using the volatility change rate. The risk confidence at the current moment is iteratively calculated by combining the risk confidence accumulated at the previous moment and the natural decay constant. When the risk confidence exceeds a preset threshold, a graded warning is executed.

2. The emergency medical resource information early warning method based on AI recognition according to claim 1, characterized in that, The step of performing initial attribute screening of the foreground moving target based on the actual volume ratio to determine the normalization adjustment coefficient includes: Calculate the real area ratio of the connected component, where the real area ratio is the ratio of the area of ​​the connected component to the area of ​​its convex hull. If the actual area ratio is lower than a preset actual area ratio threshold, it is marked as a low-risk suspected target and the value of the normalization adjustment coefficient is lowered; if it is not lower than the preset actual area ratio threshold, it is marked as a high-risk suspected target and the value of the normalization adjustment coefficient is maintained or increased.

3. The method for early warning of emergency medical resource information based on AI recognition according to claim 1, characterized in that, The calculation of the instantaneous spatial obstruction index generated by objects in the current frame using a preset logarithmic potential energy field model includes: Calculate the ratio of the potential energy radiation radius to the distance factor, add 1 to the ratio and take the natural logarithm to obtain the distance attenuation factor; wherein, the distance factor is composed of the sum of the pixel distance from the object's centroid to the emergency cabinet handle and the safety buffer constant; The instantaneous spatial obstruction index is obtained by multiplying the normalization adjustment coefficient, the pixel area of ​​the foreground connected region, and the distance attenuation factor.

4. The method for early warning of emergency medical resource information based on AI recognition according to claim 1, characterized in that, The normalization adjustment coefficient is dynamically adjusted based on the total image resolution and the morphological regularity of the foreground moving target. The morphological regularity is characterized by the real volume ratio and is used to suppress the weight of irregular objects.

5. The method for early warning of emergency medical resource information based on AI recognition according to claim 1, characterized in that, The calculation of the fluctuation rate of change based on the instantaneous spatial obstruction index of the current frame and the previous frame includes: Read the instantaneous spatial hindrance index calculated in the previous frame and the instantaneous spatial hindrance index in the current frame; The absolute value of the difference between the instantaneous spatial hindrance index of the current frame and the instantaneous spatial hindrance index of the previous frame is calculated as the fluctuation rate.

6. The method for early warning of emergency medical resource information based on AI recognition according to claim 5, characterized in that, The iterative calculation of the risk confidence level at the current moment includes: The stability weighting factor is calculated using an exponential function, where the exponential part of the stability weighting factor is the negative value of the product of the stability sensitivity coefficient and the rate of change of volatility. Multiply the instantaneous spatial obstacle index of the current frame by the stability weighting factor to obtain the risk increment of the current frame; The accumulated risk confidence from the previous moment is added to the risk increment of the current frame, and the natural decay constant is subtracted to obtain the final accumulated risk confidence at the current moment.

7. The method for early warning of emergency medical resource information based on AI recognition according to claim 6, characterized in that, The calculation of the risk confidence level also includes a boundary condition constraint step: If the calculated risk confidence score is less than zero, then the risk confidence score is forcibly assigned a value of zero; If the calculated risk confidence score is greater than the set saturation upper limit, then the risk confidence score will be forcibly assigned the value of the saturation upper limit.

8. The method for early warning of emergency medical resource information based on AI recognition according to claim 1, characterized in that, The implementation of tiered early warning includes: When the risk confidence level exceeds the first preset ratio threshold and continues to rise, a level one warning is triggered and a voice prompt is played. When the risk confidence level is greater than the second preset ratio threshold, it is determined to be a substantial obstruction, triggering a level two alarm. The current frame is captured and a heat map of the risk confidence level and the instantaneous spatial obstacle index is superimposed and pushed to the management cloud platform.

9. The method for early warning of emergency medical resource information based on AI recognition according to claim 1, characterized in that, The straight-line distance is Euclidean distance, and the center point of the operating handle is the singularity of the potential energy field, which is the area with the highest protection level.

10. An emergency medical resource information early warning system based on AI recognition, characterized in that, include: A processor and a memory, wherein the memory stores computer program instructions that, when executed by the processor, implement an AI-based emergency resource information early warning method according to any one of claims 1-9.

Images