Method for identifying ground vehicle reversing specification based on video analysis

Abstract

This invention relates to a video analysis-based method for identifying proper reversing behavior of vehicles on the apron. The method utilizes an edge computing module and a comprehensive processing module for identification. The edge computing module acquires real-time video stream data from the flight area cameras, identifies vehicle and personnel targets, structures the relevant information, and pushes it to the comprehensive processing module. The comprehensive processing module organizes the structured data, performs a comprehensive judgment, generates a violation event, and sends it to the user for review. The advantages of this invention are: after training with real airport data, it achieves high accuracy in identifying apron vehicles and personnel. This invention can identify vehicle reversing behavior and whether someone is directing it, significantly improving the depth and breadth of airport security supervision, significantly increasing the work efficiency of supervisors, and enhancing the overall airport security management level.

Description

Technical Field

[0001] This invention relates to a video analysis-based method for recognizing the proper reversing patterns of vehicles on the tarmac, in the field of airport security. Background Technology

[0002] Airport tarmac operations involve numerous vehicles across various business scenarios and complex operational relationships. Due to their involvement in flight support, safety requirements are extremely high. Airport operational regulations mandate that vehicles be directed by designated personnel during reversing to ensure safe and reliable operation. Traditional supervision relies on manual inspections by on-site supervisors to detect reversing violations. This method is inefficient and limits the number and scope of violations detected, leaving significant gaps in safety management and increasing the risk of safety issues. This invention utilizes numerous cameras deployed on the tarmac to perform real-time video flow analysis, identifying whether vehicle reversing is compliant with regulations. This significantly reduces the probability of related violations and effectively improves airport safety management. Summary of the Invention

[0003] To overcome the shortcomings of existing technologies, this invention provides a video analysis-based method for recognizing the correct reversing patterns of vehicles on the tarmac. The technical solution of this invention is as follows:

[0004] A method for identifying the correct reversing behavior of vehicles on an apron based on video analysis is proposed. This method uses an edge algorithm module and a comprehensive processing module for identification. The edge algorithm module is used to acquire real-time video stream data from the flight area camera, identify vehicle and personnel target information in it, and then process the relevant information in a structured manner and push it to the comprehensive processing module.

[0005] The integrated processing module is used to organize the target structured data, make comprehensive judgments, determine whether the vehicle is reversing, and whether there is someone directing it. If illegal reversing behavior is detected, a violation event will be generated and sent to the user terminal for the user to view.

[0006] After the edge detection algorithm module identifies the target information, it places the identified target into the vehicle dynamic list, the personnel dynamic list, and the reversing behavior analysis list. First, it traverses the vehicle dynamic list to obtain vehicle point set information. It is assumed that the polygon formed by the point set can be divided into several finite triangles T1, T2, T3, T4, ..., Tn according to the arbitrarily selected point P(px,py). The geometric centers of these triangles are denoted as TC1, TC2, TC3, TC4, ..., TCn, and their areas are denoted as TA1, TA2, TA3, TA4, ..., TAn.

[0007] The current video frame is parsed to obtain the geometric center coordinates (cx1, cy1) of the vehicle point set. When parsing the next video frame, the geometric center of the vehicle point set is recalculated and denoted as (cx2, cy2). The trigonometric function sinθ, which is the angle between the line connecting the two geometric center points of the vehicle point set and the due north direction, is used as the first reference value.

[0008] Take all points in the current vehicle point set and perform a second sinθ calculation with the geometric center (cx2, cy2) in sequence. The geometric center point (cx2, cy2) is taken as the first calculation point. All the obtained values ​​are compared with the first reference value. The point that is closest to the reference value is considered to be close to intersecting the geometric center line of the vehicle point set, and is denoted as (Ix1, Iy1).

[0009] The solution location is changed, and the points in the current vehicle point set are used as the first calculation points. The sinθ is then calculated a second time. When the sinθ of the point in the current vehicle point set that is closest to the two geometric centers is calculated, this point is considered another intersection point, denoted as (Ix2, Iy2). The distances L1 and L2 between the two intersection points and the geometric center point are then calculated.

[0010] When calculating the coordinates of a triangle:

[0011] Let the coordinates of the triangle be (x1, y1), (x2, y2), (x3, y3), and the coordinates of TCn (TCx, TCy) be calculated as follows:

[0012]

[0013] The area of ​​TAn is calculated as follows: TAn = x1y2 - x1y3 + x2y3 - x2y1 + x3y1 - x2y2;

[0014] Solve for the geometric center coordinates (cx1, cy1) of the vehicle point set:

[0015]

[0016] The calculated geometric center position is stored in the vehicle information. When parsing the next video frame, the vehicle's geometric center is recalculated, denoted as (cx2, cy2), and the angle trigonometric function sinθ is solved.

[0017] Based on the vehicle's configuration and internal structure, the direction of the vehicle's front and rear is determined by judging the values ​​of L1 and L2. If, according to the internal configuration, L1 is the largest for the rear, the vehicle is currently in reverse and is recorded in the reversing behavior analysis list. Next, personnel are identified by determining the geometric center point of the personnel (pcx1, pcy1) and calculating the sinθ value between the geometric center point of the personnel and the geometric center point of the vehicle. This sinθ value is used to determine whether personnel are directing traffic from in front of the vehicle. If the difference between this value and the sinθ value of the two geometric center points of the vehicle is less than a set constant Y, it is considered that there are personnel in front of the vehicle while it is reversing, and the reversing is considered compliant. If no one is identified, the vehicle is considered to be reversing illegally, and a vehicle reversing violation event is generated and sent to the user. The advantages of this invention are: after training with real airport data, it has high accuracy in identifying vehicles and personnel on the tarmac. This invention can identify vehicle reversing behavior and whether someone is directing traffic, significantly improving the depth and breadth of airport security supervision, significantly improving the work efficiency of supervisory personnel, and enhancing the overall security management level of the airport. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the main structure of the edge algorithm module and the integrated processing module of the present invention. Detailed Implementation

[0019] The present invention will be further described below with reference to specific embodiments, and the advantages and features of the present invention will become clearer as a result. However, these embodiments are merely exemplary and do not constitute any limitation on the scope of the present invention. Those skilled in the art should understand that modifications or substitutions can be made to the details and form of the technical solutions of the present invention without departing from the spirit and scope of the present invention, but all such modifications and substitutions fall within the protection scope of the present invention.

[0020] See Figure 1 This invention relates to a video analysis-based method for identifying the proper reversing behavior of vehicles on the tarmac. The method utilizes an edge computing module and a comprehensive processing module for identification. The edge computing module acquires real-time video stream data from the flight area cameras, identifies vehicle and personnel targets within the stream, and structures and processes this information before pushing it to the comprehensive processing module. The comprehensive processing module organizes the structured data, performs a comprehensive judgment, and determines whether the vehicle is reversing and whether anyone is directing it. If a violation is detected, a violation event is generated and sent to the user for review.

[0021] After the edge detection module identifies the target information, it places the identified target into the vehicle dynamic list, the personnel dynamic list, and the reversing behavior analysis list. First, it traverses the vehicle dynamic list to obtain the vehicle point set information. It is assumed that the polygon formed by the point set can be divided into several finite triangles T1, T2, T3, T4, ..., Tn according to the arbitrarily selected point P(px,py). The geometric centers of these triangles are denoted as TC1, TC2, TC3, TC4, ..., TCn, and their areas are denoted as TA1, TA2, TA3, TA4, ..., TAn.

[0022] The current video frame is parsed to obtain the geometric center coordinates (cx1, cy1) of the vehicle point set. When parsing the next video frame, the geometric center of the vehicle point set is recalculated and denoted as (cx2, cy2). The trigonometric function sinθ, which is the angle between the line connecting the two geometric center points of the vehicle point set and the due north direction, is used as the first reference value.

[0023] Take all points in the current vehicle point set and perform a second sinθ calculation with the geometric center (cx2, cy2) in sequence. The geometric center point (cx2, cy2) is taken as the first calculation point. All the obtained values ​​are compared with the first reference value. The point that is closest to the reference value is considered to be close to intersecting the geometric center line of the vehicle point set, and is denoted as (Ix1, Iy1).

[0024] The solution location is changed, and the points in the current vehicle point set are used as the first calculation points. The sinθ is then calculated a second time. When the sinθ of the point in the current vehicle point set that is closest to the two geometric centers is calculated, this point is considered another intersection point, denoted as (Ix2, Iy2). The distances L1 and L2 between the two intersection points and the geometric center point are then calculated.

[0025] When calculating the coordinates of a triangle:

[0026] Let the coordinates of the triangle be (x1, y1), (x2, y2), (x3, y3), and the coordinates of TCn (TCx, TCy) be calculated as follows:

[0027]

[0028] The area of ​​TAn is calculated as follows: TAn = x1y2 - x1y3 + x2y3 - x2y1 + x3y1 - x2y2;

[0029] Solve for the geometric center coordinates (cx1, cy1) of the vehicle point set:

[0030]

[0031]

[0032] The calculated geometric center position is stored in the vehicle information. When parsing the next video frame, the vehicle's geometric center is recalculated, denoted as (cx2, cy2), and the trigonometric function sinθ, representing the angle between the line connecting the two center points and the due north direction, is calculated.

[0033] Take all points in the current vehicle point set and calculate sinθ sequentially with the geometric center (cx2, cy2). The geometric center (cx2, cy2) is taken as the first calculation point. When this value is closest to the sinθ values ​​of the two geometric centers, the point is considered to be close to intersecting the two geometric center lines, denoted as (Ix1, Iy1). Change the solution position, taking the points in the current vehicle point set as the first calculation point, and solve for sinθ. When the sinθ value is closest to the two geometric centers, this point is considered another intersection point, denoted as (Ix2, Iy2). Calculate the distances L1 and L2 between the two intersection points and the geometric center point.

[0034] Based on the vehicle's configuration and internal structure, the direction of the vehicle's front and rear is determined by judging the values ​​of L1 and L2. If, according to the internal configuration, L1 is the largest and the rear is the largest, then the vehicle is currently in reverse and is recorded in the reversing behavior analysis list. Next, the personnel are identified, and the geometric center point of the personnel (pcx1, pcy1) is calculated. The sinθ value between the geometric center point of the personnel and the geometric center point of the vehicle is calculated to determine whether the personnel are in front of the vehicle giving directions. When the difference between this value and the sinθ value of the two geometric center points of the vehicle is less than the set constant Y, it is considered that there are personnel in front of the vehicle while it is reversing, and the vehicle is considered to be reversing legally. When no one is identified, the vehicle is considered to be reversing illegally, and a vehicle reversing violation event is generated and sent to the user.

[0035] In this invention, sinθ is calculated three times to find the point where the extension of the line connecting the two center points intersects the vehicle point set. Since the extension line intersects the vehicle point set polygon at two points, the angle sinθ is calculated twice more based on the angle of the two center points to compare the two points that are closest to the extension of the line connecting the two center points.

[0036] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for recognizing the correct reversing posture of vehicles on an apron based on video analysis, characterized in that, The method for identifying standard reversing vehicles on the apron is achieved through the use of an edge algorithm module and a comprehensive processing module. The edge algorithm module is used to acquire real-time video stream data from the flight area camera, identify vehicle and personnel target information, and then process the relevant information in a structured manner and push it to the comprehensive processing module. The integrated processing module is used to organize the target structured data, make comprehensive judgments, determine whether the vehicle is reversing, and whether there is someone directing it. If illegal reversing behavior is detected, a violation event will be generated and sent to the user terminal for the user to view. After the edge detection algorithm module identifies the target information, it places the identified target into the vehicle dynamic list, the personnel dynamic list, and the reversing behavior analysis list. First, it traverses the vehicle dynamic list to obtain the vehicle point set information. It is assumed that the polygon formed by the point set can be divided into several finite triangles T1, T2, T3, T4, ..., Tn according to the arbitrarily selected point P (px, py). The geometric centers of these triangles are denoted as TC1, TC2, TC3, TC4, ..., TCn, and their areas are denoted as TA1, TA2, TA3, TA4, ..., TAn. The current video frame is parsed to obtain the geometric center coordinates of the vehicle point set (cx1, cy1); when parsing the next video frame, the geometric center of the vehicle point set is recalculated and denoted as (cx2, cy2). The trigonometric function sinθ, which is the angle between the line connecting the two geometric center points of the vehicle point set and the due north direction, is used as the first reference value. Take all points in the current vehicle point set and perform a second sinθ calculation with the geometric center (cx2, cy2) in sequence. The geometric center point (cx2, cy2) is taken as the first calculation point. All the obtained values ​​are compared with the first reference value. The point that is closest to the reference value is considered to be close to intersecting the geometric center line of the vehicle point set, and is denoted as (Ix1, Iy1). The solution location is changed, and the points in the current vehicle point set are used as the first calculation points. The sinθ is then calculated a second time. When the sinθ of the point in the current vehicle point set that is closest to the two geometric centers is calculated, this point is considered another intersection point, denoted as (Ix2, Iy2). The distances L1 and L2 between the two intersection points and the geometric center point are then calculated. , .

2. The method for identifying the correct reversing procedure of apron vehicles based on video analysis according to claim 1, characterized in that, When calculating the coordinates of a triangle: Let the coordinates of the triangle be (x1, y1), (x2, y2), (x3, y3), and the coordinates of TCn (TCx, TCy) be calculated as follows: ； The area of ​​TAn is calculated as follows: ; Solve for the geometric center coordinates (cx1, cy1) of the vehicle point set: ； The calculated geometric center position is stored in the vehicle information. When parsing the next video frame, the vehicle's geometric center is recalculated, denoted as (cx2, cy2), and the angle trigonometric function sinθ is solved. .

3. The method for identifying the correct reversing pattern of apron vehicles based on video analysis according to claim 2, characterized in that, Based on the vehicle's configuration and internal structure, the direction of the vehicle's front and rear is determined by judging the values ​​of L1 and L2. If, according to the internal configuration, L1 is the largest and the rear is the largest, then the vehicle is currently in a reversing state, and this vehicle is recorded in the reversing behavior analysis list. Next, the personnel are identified, and the geometric center point of the personnel (pcx1, pcy1) is calculated. The sinθ value between the geometric center point of the personnel and the geometric center point of the vehicle is calculated to determine whether the personnel are giving directions in front of the vehicle. When the difference between this value and the sinθ value of the two geometric center points of the vehicle is less than the set constant Y, it is considered that there are personnel in front of the vehicle while it is reversing, and the vehicle is considered to be reversing legally. When it is determined that there are no personnel, the vehicle is considered to be reversing illegally, and a vehicle reversing violation event will be generated and sent to the user terminal.

Images