Security protection system and method based on space calculation and computer vision detection

Abstract

The invention discloses a security escort system and method based on space calculation and computer vision detection. The method comprises the following steps: S1, acquiring escort vehicle information and escort route information in real time; s2, dynamically matching the acquired escorted vehicle information with the escorted route information, and monitoring the driving path of the escorted vehicle in real time; s3, performing spatial matching on the escorted vehicle information and the escorted route information to dynamically display the real-time position of the escorted vehicle; s4, judging whether an abnormal road condition exists in the escort route, if so, generating early warning information, and carrying out notification early warning, otherwise, continuing monitoring; and S5, judging whether an emergency event exists in the escort route or not, if so, calculating the event distance and the risk level, generating emergency early warning information, and carrying out notification early warning, otherwise, continuing monitoring until the escort task is finished, dynamically matching the position of the escort vehicle with the escort road condition, rapidly identifying the emergency event, and carrying out early warning. And the security protection task is ensured to be safely and smoothly carried out.

Description

Technical Field

[0001] The present invention relates to the technical field related to traffic safety, and in particular to a security navigation system and method based on spatial computing and computer vision detection. Background Art

[0002] In the field of traffic safety, intelligent dynamic monitoring and emergency warning are of great significance to the smooth implementation of escort missions. In the existing technology, GPS or cameras are usually used alone for vehicle position monitoring or road monitoring, which cannot achieve efficient linkage between the two and is difficult to meet the high standards of security and navigation for global monitoring and real-time warning. Superimposing the GPS position and camera position for spatial calculation, and applying computer vision to real-time video detection to analyze abnormal road conditions and emergencies in the vehicle's surrounding environment can provide higher response speed and accuracy. Especially in escort scenarios, computer vision can analyze video images in real time, automatically identify abnormal behaviors (such as congestion, obstacles) and emergencies (such as accidents, pedestrian intrusions), realize full-process monitoring and early warning, and ensure the safe passage of escort vehicles.

[0003] However, the existing technology mainly uses GPS to track the vehicle position and uses cameras to monitor road conditions, but the two lack effective linkage. GPS mainly relies on satellite signals, while camera monitoring is fixed-point monitoring. It is impossible to track the vehicle in real time and provide real-time warnings for emergencies, that is, there are the following problems: 1. GPS and camera data are independent of each other, making it difficult to achieve dynamic data matching and unable to conduct joint analysis of the real-time location of a specific vehicle and the road conditions ahead; 2. Lack of real-time perception and prediction capabilities ahead of the escort route, unable to detect emergencies in a timely manner, unable to implement route planning and proactive response to emergencies; 3. It is impossible to achieve effective early warning and ensure the safety of the entire security and navigation mission. Summary of the invention

[0004] The purpose of the present invention is to provide a security navigation system and method based on spatial computing and computer vision detection to solve the problems raised in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides the following technical solution: a security navigation method based on spatial computing and computer vision detection, the method comprising the following steps: S1. Real-time collection of escort vehicle information and escort route information; S2. Dynamically match the collected escort vehicle information with the escort route information, and monitor the driving path of the escort vehicle in real time; S3, spatially matching the escort vehicle information and the escort route information to dynamically display the real-time position of the escort vehicle; S4. Determine whether there are abnormal road conditions on the escort route. If so, generate warning information and issue a notification warning. Otherwise, continue monitoring. S5. Determine whether there is an emergency on the escort route. If so, calculate the event distance and risk level, generate emergency warning information, and issue a notification warning. Otherwise, continue monitoring until the escort mission is completed.

[0006] Preferably, the escort vehicle information in step S1 includes the longitude and latitude, speed and direction information of the escort vehicle, and the escort route information includes the escort route video stream information and the geographic location information of the acquisition device.

[0007] Preferably, the dynamic matching in step S2 includes the following sub-steps: S21. Modeling the collection equipment and geographic information of the escort roads; S22, acquiring escort vehicle information in real time; S23, constructing a spatial query range with the escort vehicle as the center through a spatial algorithm, wherein the area in front of the escort vehicle is a polygonal area, and the area behind the escort vehicle is a rectangular area; S24. Sort the escort road collection devices based on the front and rear distances of the escort vehicles to achieve dynamic matching between the escort vehicles and the escort roads.

[0008] Preferably, the abnormal road conditions in step S4 include traffic congestion, obstacles, road construction, waterlogging and bad weather.

[0009] Preferably, the emergency in step S5 includes a traffic accident, a pedestrian or a non-motor vehicle entering the escort road.

[0010] Preferably, the calculation of the emergency distance and risk level in step S5 includes the following sub-steps: S51. Assign different weights according to the impact of different types of events on the escort mission; S52, extracting key features of the emergency event from the escort road information; S53. Assign scores to the severity of the event through multi-dimensional scoring and obtain the total risk score Q , the formula is as follows: T : Event type weight score; D : The distance score between the event and the vehicle; A : Event coverage area score; E : Correction value of environmental factors; P : Event duration score; W 1, W 2,W 3. W 4. W 5: The weight of each indicator; S54. Classify the severity of the emergency according to the total risk score and send corresponding warning information.

[0011] Preferably, the key features of the emergency event in step S52 include event type, event location and dynamic information of the event.

[0012] The present invention also discloses a security and navigation system based on spatial computing and computer vision detection, including a data acquisition module and a data processing module. The data acquisition module includes a vehicle-mounted GPS module and a camera module. The vehicle-mounted GPS module is installed on the escort vehicle and is used to collect escort vehicle information in real time. The camera module is arranged along the key sections of the escort route and is used to collect escort route information in real time.

[0013] Preferably, the key sections include areas with complex traffic, high-risk sections, strategic locations that escort vehicles must pass through during the mission, traffic control areas, and escort vehicle mission starting and ending areas.

[0014] Preferably, the data processing module comprises: Data fusion module: spatially matches the collected escort vehicle information and escort route information to dynamically display the real-time position of the escort vehicle on the escort route; Abnormal road condition detection module: The algorithm is used to analyze abnormal conditions in the escort route information in real time. If an abnormal event is detected, it will be combined with the real-time position of the escort vehicle to generate corresponding warning information; Emergency Identification Module: used to identify emergencies on the escort route of the escort vehicle, evaluate the severity of the emergencies, and generate effective emergency warnings based on the escort vehicle information; Real-time response and warning module: informs the driver and the command center of detected abnormal road conditions and / or emergency information in visual and / or auditory form. Beneficial effects of the present invention: 1. The present invention proposes a spatial calculation method based on the vehicle GPS and camera position to achieve dynamic matching monitoring of the vehicle position and road conditions along the way. At the same time, combined with computer vision detection algorithms, real-time detection of data from the front and rear cameras of the vehicle is performed to quickly analyze and identify emergencies and issue early warnings, thereby ensuring the safe and smooth progress of security and navigation tasks.

[0015] 2. The present invention automatically identifies abnormal road conditions and emergencies and combines computer vision technology to enable the system to automatically identify abnormal behaviors and emergencies in the video and obtain real-time information on the road conditions and emergencies ahead.

[0016] 3. The present invention combines the vehicle's GPS position to timely warn of potential risks, ensuring that sufficient reaction time is provided for escort vehicles before an incident occurs, achieving accurate warnings, and improving the response speed to escort vehicles.

[0017] 4. The present invention effectively ensures the safe driving of escort vehicles in emergency situations and the smooth progress of security and escort missions through real-time detection and analysis and rapid early warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A schematic diagram of the steps of the secure navigation method according to an embodiment of the present invention; Figure 2 Schematic diagram of data processing flow of each module of the system in an embodiment of the present invention. DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0020] See also Figure 1 and Figure 2 The present invention provides a security navigation method based on spatial computing and computer vision detection, the method comprising the following steps: S1. Real-time collection of escort vehicle information and escort route information; S2. Dynamically match the collected escort vehicle information with the escort route information, and monitor the driving path of the escort vehicle in real time; S3, spatially matching the escort vehicle information and the escort route information to dynamically display the real-time position of the escort vehicle; S4. Determine whether there are abnormal road conditions on the escort route. If so, generate warning information and issue a notification warning. Otherwise, continue monitoring. S5. Determine whether there is an emergency on the escort route. If so, calculate the event distance and risk level, generate emergency warning information, and issue a notification warning. Otherwise, continue monitoring until the escort mission is completed.

[0021] Preferably, the escort vehicle information in step S1 includes the longitude and latitude, speed and direction information of the escort vehicle, and the escort route information includes the escort route video stream information and the geographic location information of the acquisition device.

[0022] Preferably, the dynamic matching in step S2 includes the following sub-steps: S21. Modeling the collection equipment and geographic information of the escort roads; S22, acquiring escort vehicle information in real time; S23, constructing a spatial query range with the escort vehicle as the center through a spatial algorithm, wherein the area in front of the escort vehicle is a polygonal area, and the area behind the escort vehicle is a rectangular area; S24. Sort the escort road collection devices based on the front and rear distances of the escort vehicles to achieve dynamic matching between the escort vehicles and the escort roads.

[0023] Preferably, the abnormal road conditions in step S4 include traffic congestion, obstacles, road construction, waterlogging and bad weather.

[0024] Preferably, the emergency in step S5 includes a traffic accident, a pedestrian or a non-motor vehicle entering the escort road.

[0025] Preferably, the calculation of the emergency distance and risk level in step S5 includes the following sub-steps: S51. Assign different weights according to the impact of different types of events on the escort mission; S52, extracting key features of the emergency event from the escort road information; S53. Assign scores to the severity of the event through multi-dimensional scoring and obtain the total risk score Q , the formula is as follows: T : Event type weight score; D : The distance score between the event and the vehicle; A : Event coverage area score; E : Correction value of environmental factors; P : Event duration score; W 1, W 2, W 3. W 4. W 5: The weight of each indicator; S54. Classify the severity of the emergency according to the total risk score and send corresponding warning information.

[0026] Preferably, the key features of the emergency event in step S52 include event type, event location and dynamic information of the event.

[0027] The present invention also discloses a security and navigation system based on spatial computing and computer vision detection, including a data acquisition module and a data processing module. The data acquisition module includes a vehicle-mounted GPS module and a camera module. The vehicle-mounted GPS module is installed on the escort vehicle and is used to collect escort vehicle information in real time. The camera module is arranged along the key sections of the escort route and is used to collect escort route information in real time.

[0028] Preferably, the key sections include areas with complex traffic, high-risk sections, strategic locations that escort vehicles must pass through during the mission, traffic control areas, and escort vehicle mission starting and ending areas.

[0029] Preferably, the data processing module comprises: Data fusion module: spatially matches the collected escort vehicle information and escort route information to dynamically display the real-time position of the escort vehicle on the escort route; Abnormal road condition detection module: The algorithm is used to analyze abnormal conditions in the escort route information in real time. If an abnormal event is detected, it will be combined with the real-time position of the escort vehicle to generate corresponding warning information; Emergency Identification Module: used to identify emergencies on the escort route of the escort vehicle, evaluate the severity of the emergencies, and generate effective emergency warnings based on the escort vehicle information; Real-time response and warning module: informs the driver and the command center of detected abnormal road conditions and / or emergency information in visual and / or auditory form.

[0030] The present invention proposes an intelligent escort system based on vehicle GPS and camera position, which uses multiple modules to work together to achieve full-time real-time monitoring and intelligent early warning of escort vehicles. The escort system consists of a data acquisition module and a data processing module, which are as follows: 1. The data acquisition module includes vehicle GPS module and camera module: 1. The vehicle GPS module includes a GPS positioning device, a data acquisition unit and a wireless transmission module.

[0031] The vehicle GPS module is installed on the escort vehicle and is mainly responsible for real-time collection of the vehicle's location information, including latitude and longitude, speed and driving direction. Through a high-precision positioning system and data upload interface, the vehicle GPS module ensures that the escort vehicle's location information can be updated in real time and transmitted to the data processing module during the movement of the escort vehicle.

[0032] The vehicle-mounted GPS module receives satellite signals to calculate the real-time latitude and longitude, speed and direction information of the escort vehicle, and uploads the data to the data fusion module through the wireless transmission module. The real-time location of the escort vehicle ensures the accuracy and timeliness of the escort vehicle location data.

[0033] 2. The camera module includes a high-resolution camera, image acquisition and transmission module.

[0034] Camera modules are deployed along key sections of the escort route to monitor the escort route and its surrounding environment. Each camera can collect video streams in real time and include its own geographic location information. The key sections of the escort route refer to the sections that need to be monitored and prevented during the escort process.

[0035] The selection of key sections of the escort road is based on the needs of the escort mission, the characteristics of the actual road conditions, and factors that may pose risks to the escort vehicles. Key sections include the following types: 1. Complex traffic areas with large traffic volume, dense intersections, and prone to congestion; 2. High-risk sections with frequent traffic accidents and emergencies in history or high risks; 3. Strategic locations that vehicles must pass through during the escort mission, including important landmarks and nodes such as entrances and exits, checkpoints, and escort mission target areas; 4. Traffic control areas affected by restricted traffic, speed limits, or temporary traffic control; 5. The starting position of the escort vehicle and the mission end area.

[0036] By installing camera modules at key road sections, the cameras capture the road condition information of the escort route in real time, and transmit the video stream data together with the camera's geographic location data to the data fusion module. The data from the camera modules provides a rich data source for subsequent abnormal road condition detection and emergency event identification.

[0037] 2. The data processing module includes the vehicle and camera matching algorithm module, data fusion module, abnormal road condition detection module, emergency event recognition module and real-time response and warning module: 3. Vehicle and camera matching algorithm module The vehicle and camera matching algorithm module dynamically calculates the information of the M cameras in front of the escort vehicle and the N cameras behind the escort vehicle according to the real-time GPS position information of the escort vehicle, so as to monitor the driving path of the escort vehicle. The specific implementation method of the camera and vehicle position matching algorithm based on spatial calculation is as follows: 1) Geographic information modeling of cameras and roads a. Camera geographic information: Collect and store in advance the geographic location information (latitude and longitude) and coverage (such as angle and radius) of all cameras deployed along the escort route.

[0038] b. Road segment modeling: The escort route is divided into several small segments through manual cutting and segmentation by the user on the page, and the starting point, end point and direction angle of each segment are defined to quickly match the geometric relationship between the camera and the route.

[0039] c. Data storage form: Camera information and road segment information are stored in a spatial database (such as PostGIS) for fast query.

[0040] 2) Acquisition of real-time escort vehicle location a. Vehicle GPS data: The GPS module of the escort vehicle uploads the current location in real time, including latitude and longitude, driving direction, speed, etc.

[0041] b. Data update frequency: Based on the escort mission requirements, the frequency of GPS data upload is usually every second or shorter.

[0042] 3) Spatial range screening a. Construct a range centered on the vehicle: Based on the current position (Lat, Lon) of the vehicle and its travel direction, a spatial query range is constructed using a spatial database such as PostGIS. For example, the ST_Buffer and ST_Rotate functions are used to generate an area 500 meters in front of and 200 meters behind the escort vehicle, respectively, and finally an irregular polygon is obtained.

[0043] b. Front range: Use GIS tools to create a polygonal area extending in the direction of the vehicle, covering the road section that the escort vehicle is about to arrive at. The specific steps are as follows: (1) The longitude and latitude (Lat, Lon) of the escort vehicle’s current location, the vehicle’s direction of travel (azimuth angle θ), and the query range (e.g., 500 meters ahead, 50 meters to the left and right) are used as input; (2) Then, a center line with a length of L is extended according to the vehicle’s current position (Lat, Lon) and direction angle θ through the PostGIS function.

[0044] (3) Use the center line to apply the left and right buffer distances and use the ST_Buffer function to generate a rectangular area. If a more complex front range (such as an arc area) is required, the shape parameters of ST_Buffer can be adjusted conventionally.

[0045] (4) The geometric data of the buffer zone is converted into a polygon through the ST_AsPolygon function to form the coverage area in front of the vehicle.

[0046] c. Rear range: Similarly, a rectangular area at a certain distance behind the escort vehicle is realized.

[0047] d. Query camera information: Filter the location information of all cameras within the range in the database.

[0048] 4) Distance-based sorting a. Sort the selected cameras by their distance from the vehicle: b. M cameras in front: Starting from the current position of the escort vehicle, select the cameras that cover the vehicle's forward path in order of distance from near to far.

[0049] c. N rear cameras: Starting from the current position of the escort vehicle, select the cameras that cover the path behind the vehicle.

[0050] 5) Dynamic update and cache optimization a. Regularly refresh matching results: As the vehicle moves, the algorithm will periodically recalculate the matching results and update the camera lists in front and behind.

[0051] b. Cache optimization: For camera matching results that do not change in a short period of time, a cache mechanism is used to reduce the calculation frequency, and the full matching calculation is triggered only when the vehicle moves to a new road section.

[0052] 6) Implement technology and algorithm support Spatial computing framework: Use spatial databases (such as PostGIS) or GIS libraries (such as GeoTools) to perform efficient range queries and geometric calculations.

[0053] The matching algorithm is used to perform spatial calculations on the GPS location information of the escort vehicle, so as to obtain real-time cameras within a certain range in front of and behind the escort vehicle. M front cameras cover the area that the vehicle is about to arrive at, and N rear cameras provide monitoring of the area behind the vehicle. By dynamically matching and selecting cameras, potential risks on the escort route can be detected in time and early warnings can be provided.

[0054] 4. The data fusion module includes data synchronization module, spatial fusion module and visualization module.

[0055] The data fusion module is responsible for spatially matching the escort vehicle's GPS information and camera video data to dynamically display the escort vehicle's real-time position on the scheduled escort route. The data fusion module receives data from the vehicle's GPS module and camera module, uses spatial calculations to dynamically match the escort vehicle's position with the collected video stream data, and displays the vehicle's position and surrounding real-time road conditions on the monitoring interface through the visualization module, so that the command center and the driver can intuitively see the escort vehicle's position on the road and the real-time status of the area it is about to arrive at.

[0056] 5. The abnormal road condition detection module includes a computer vision processing unit and an anomaly detection algorithm library.

[0057] The abnormal road condition detection module uses computer vision technology to perform real-time analysis of abnormal situations in the camera video stream. The module identifies abnormal situations such as traffic congestion, obstacles, water accumulation and bad weather through object detection and behavior analysis algorithms.

[0058] The camera module transmits the collected video stream data to the abnormal road condition detection module, and the computer vision algorithm automatically identifies the type of abnormal road conditions (such as obstacles or congestion). The detected abnormal events are combined with the real-time position of the escort vehicle to generate corresponding warning information, providing timely safety reminders for the driver.

[0059] 6. The emergency event recognition module includes a deep learning processing unit and an event recognition model library.

[0060] The emergency recognition module is used to identify emergencies such as traffic accidents, pedestrians or non-motor vehicles intruding into the escort vehicle's lane, etc. This module can evaluate the severity of the incident and generate effective warnings based on the vehicle's location information. By inputting the image data in the camera video stream into the emergency recognition module, the system identifies the type of emergency (such as traffic accidents) through a deep learning model. After identifying the emergency, the system calculates the distance between the incident and the escort vehicle and evaluates the risk level of the incident. If the incident poses a potential threat to the escort vehicle, the system will issue an emergency warning to remind the driver to avoid it in time.

[0061] The emergency event recognition module uses analysis technology based on deep learning algorithms to comprehensively evaluate the severity of the event based on multiple dimensions such as event type, spatial relationship between event location and vehicle, and environmental characteristics. The specific methods are as follows: 1. Evaluation indicators of the severity of the incident. For different types of emergencies, a set of common evaluation indicators are defined, including the type of incident, the distance between the location of the incident and the escort vehicle, the size of the area affected by the incident, and the duration of the incident, so as to quantify the risk of the incident and its potential impact on the escort mission.

[0062] a. Weights are assigned based on the different impacts of different event types on the escort mission. For example, traffic accidents: high weight, which may cause the route to be completely blocked; pedestrians entering the lane: medium weight, which may cause vehicles to stop or avoid for a short time; non-motor vehicles entering the lane: low weight, which usually only requires slight deceleration to avoid; The event classification model is based on the deep learning algorithm YOLO or Mask R-CNN to classify events in the video stream collected by the camera.

[0063] b. The distance between the location where the incident occurred and the vehicle. The closer the distance, the higher the risk and the greater the weight. The Euclidean distance or driving path distance between the two is calculated using the data uploaded by the vehicle's GPS module and the geographic coordinates of the location of the incident, and is dynamically classified according to the distance: high risk: within 50 meters; medium risk: between 50-200 meters; low risk: more than 200 meters away.

[0064] c. The size of the area affected by the event. The event coverage area (such as the number of lanes involved in the traffic accident and the length of the blockage) is estimated through the event classification algorithm and the camera data stream. The larger the area, the more significant the impact on vehicle traffic and the higher the weight.

[0065] d. Duration of the event. The longer the duration, the more significant the impact of the event on the vehicle. If the event appears continuously in multiple frames of video, the risk weight will increase.

[0066] 2. Evaluation process and algorithm implementation (1) In the event recognition stage, the camera video stream is analyzed in real time through a deep learning model to identify the type and location of the event. Typical choices of deep learning models include the YOLO series, which is suitable for real-time detection and fast recognition speed, or Mask R-CNN, which is suitable for high-precision scenarios and can segment event areas.

[0067] (2) Event feature extraction: extracting key features of the event from the video, including: event type (accident, pedestrian intrusion, etc.); event location (geographic coordinates, coverage area); and dynamic information of the event (whether it is continuing).

[0068] (3) Event risk scoring, using a multidimensional scoring method to assign scores to the severity of the event; formula example: T : Event type weight score; D : The distance score between the event and the vehicle; A : Event coverage area score; E : Correction value of environmental factors; P : Event duration score; W 1, W 2, W 3. W 4. W 5 refers to the weight of each indicator, which is set according to actual needs.

[0069] (4) The incidents are divided into three levels according to the total risk score: high risk (above 80 points): an emergency warning must be issued immediately and the vehicle is advised to slow down or change route; medium risk (50-80 points): the driver is reminded to observe carefully and slow down in advance; low risk (below 50 points): the incident information is recorded and no additional measures are required.

[0070] 7. The real-time response and warning module includes a warning information generation unit, a voice prompt module and an image pop-up module.

[0071] The real-time response and warning module is mainly responsible for informing the driver and the command center of the detected abnormalities and emergency information in a visual or auditory manner.

[0072] The early warning system receives the early warning information transmitted by the abnormal road condition detection and emergency event identification modules, and generates response prompts based on the type and severity of the event. The early warning system supports voice prompts and image pop-ups. Voice prompts can directly remind drivers to pay attention to safety, and image pop-ups display event details and locations. In addition, the early warning information can also be transmitted synchronously to the command center, allowing relevant personnel to track the safety status of escort vehicles in real time and take necessary response measures.

[0073] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modification, equivalent substitution or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A security navigation method based on spatial computing and computer vision detection, characterized in that: The method comprises the following steps: S1. Real-time collection of escort vehicle information and escort route information; S2. Dynamically match the collected escort vehicle information with the escort route information, and monitor the driving path of the escort vehicle in real time; S3, spatially matching the escort vehicle information and the escort route information to dynamically display the real-time position of the escort vehicle; S4. Determine whether there are abnormal road conditions on the escort route. If so, generate warning information and issue a notification warning. Otherwise, continue monitoring. S5. Determine whether there is an emergency on the escort route. If so, calculate the event distance and risk level, generate emergency warning information, and issue a notification warning. Otherwise, continue monitoring until the escort mission is completed.

2. A security navigation method based on spatial computing and computer vision detection according to claim 1, characterized in that: The escort vehicle information in step S1 includes the longitude and latitude, speed and direction information of the escort vehicle, and the escort route information includes the escort route video stream information and the geographic location information of the acquisition device.

3. The security navigation method based on spatial computing and computer vision detection according to claim 1 is characterized in that: The dynamic matching in step S2 includes the following sub-steps: S21. Modeling the collection equipment and geographic information of the escort roads; S22, acquiring escort vehicle information in real time; S23, constructing a spatial query range with the escort vehicle as the center through a spatial algorithm, where the area in front of the escort vehicle is a polygonal area, and the area behind the escort vehicle is a rectangular area; S24. Sort the escort road collection devices based on the front and rear distances of the escort vehicles to achieve dynamic matching between the escort vehicles and the escort roads.

4. The security navigation method based on spatial computing and computer vision detection according to claim 1 is characterized in that: The abnormal road conditions in step S4 include traffic congestion, obstacles, road construction, waterlogging and bad weather.

5. The security navigation method based on spatial computing and computer vision detection according to claim 1 is characterized in that: The emergencies in step S5 include traffic accidents, pedestrians or non-motor vehicles entering the escort road.

6. The security navigation method based on spatial computing and computer vision detection according to claim 1 is characterized in that: The calculation of the emergency distance and risk level in step S5 includes the following sub-steps: S51. Assign different weights according to the impact of different types of events on the escort mission; S52, extracting key features of the emergency event from the escort road information; S53. Allocate scores to the severity of the event through multi-dimensional scoring, and obtain the total risk score Q. The formula is as follows: T: event type weight score; D: distance score between the event and the vehicle; A: event coverage area score; E: Environmental factor correction value; P: event duration score; W1, W2, W3, W4, W5: weights of each indicator; S54. Classify the severity of the emergency according to the total risk score and send corresponding warning information.

7. A security navigation method based on spatial computing and computer vision detection according to claim 6, characterized in that: The key features of the emergency event in step S52 include the event type, event location and dynamic information of the event.

8. A security and navigation system according to any one of claims 1 to 7, characterized in that: It includes a data acquisition module and a data processing module. The data acquisition module includes a vehicle-mounted GPS module and a camera module. The vehicle-mounted GPS module is installed on the escort vehicle and is used to collect escort vehicle information in real time. The camera module is arranged along the key sections of the escort route and is used to collect escort route information in real time.

9. The security navigation method based on spatial computing and computer vision detection according to claim 8, characterized in that: The key sections include areas with complex traffic, high-risk sections, strategic locations that escort vehicles must pass through during the mission, traffic control areas, and the starting and ending areas of the escort vehicle mission.

10. A security navigation method based on spatial computing and computer vision detection according to claim 9, characterized in that: The data processing module comprises: Data fusion module: spatially matches the collected escort vehicle information and escort route information to dynamically display the real-time position of the escort vehicle on the escort route; Abnormal road condition detection module: The algorithm is used to analyze abnormal conditions in the escort route information in real time. If an abnormal event is detected, it will be combined with the real-time position of the escort vehicle to generate corresponding warning information; Emergency Identification Module: used to identify emergencies on the escort route of the escort vehicle, evaluate the severity of the emergencies, and generate effective emergency warnings based on the escort vehicle information; Real-time response and warning module: informs the driver and the command center of detected abnormal road conditions and / or emergency information in visual and / or auditory form.

Images