A traffic hub passenger luggage tail swing state identification and safety monitoring method

Abstract

This invention proposes a method for identifying and monitoring the tail-wagging state of passenger luggage in transportation hubs. The method includes the following steps: setting up monitoring cameras and loudspeaker alerts in densely populated areas of the transportation hub; activating the monitoring cameras to capture video frames and providing them to a monitoring processor; the monitoring processor calculating the angular force, rate of change of center of mass distance, and angular velocity of the luggage in the monitored area using single-target tracking algorithms MMTrack and findContours, respectively; and detecting the tail-wagging state of the luggage based on a luggage tail-wagging state identification model. This invention identifies the motion characteristics of the luggage throughout its entire motion and can predict its subsequent motion state. Based on the identification results, timely warnings can be issued, effectively avoiding potential hazards caused by luggage tail-wagging, which is of great significance for safety management in densely populated transportation hubs.

Description

Technical Field

[0001] This invention relates to a passenger luggage tail-swing state recognition model and application that takes into account combined motions, and is applicable to densely populated transportation hubs. Background Technology

[0002] The problem of luggage tailing during passenger movement has always been a challenge in crowd stability analysis at transportation hubs. In densely populated public places, it often leads to chaos and disorder, and may cause serious safety accidents such as stampedes. Currently, although some studies have focused on the relationship between luggage and overall crowd evacuation, research on the identification and detection of luggage tailing remains relatively limited. Luggage tailing can cause harm to passengers and those around them; therefore, there is a need to develop a technology for identifying luggage tailing to improve travel safety.

[0003] Currently, research on luggage tail-swing recognition still has several shortcomings: 1) Some studies have focused on methods for detecting luggage lifting, such as Baumgartner's T [1] Different colors are used to represent the movement relationship between passengers and their belongings, but this research hasn't specifically focused on identifying luggage tailing states; it's merely categorizing passengers and suitcases. 2) Current research on luggage and passenger movement mainly focuses on the macroscopic impact of passengers carrying luggage on overall crowd dispersal. For example, Kitazono Y et al. proposed a method to detect luggage lifting, which can monitor abnormal behavior such as passengers discarding luggage and issue warnings. However, research on the microscopic tailing state of luggage is lacking. 3) The tailing state of luggage is closely related to the movement state of passengers. Current research on abnormal luggage state identification lacks in-depth research on the movement relationship between luggage and passengers. Summary of the Invention

[0004] To further address the problem of luggage tail-swing state recognition, this invention proposes a novel luggage tail-swing state recognition model that considers combined motions. When applied, it employs new technical means such as kinematic analysis and machine learning algorithms, proposing a more comprehensive and accurate motion state detection method, making the recognition of luggage tail-swing state more accurate and faster.

[0005] The technical solution of this invention is as follows:

[0006] A model for recognizing the tail-swing state of passenger luggage in transportation hubs is characterized by the following determination of the tail-swing effect of luggage:

[0007]

[0008] in:

[0009] 1. The angular force F acting on the suitcase starting from frame t. 转,t The maximum angular force F of small-amplitude rotation during the motion process is greater than that of the maximum rotation force. 转，max ;

[0010] 2. At frame t, the rate of change ΔT of the centroid distance between the suitcase and the passenger. a,t Greater than 0.2 m / s, but less than 0.5 m / s;

[0011] 3. At frame t, the angular velocity of the movement between the suitcase and the passenger. Greater than the maximum angular velocity of normal motion

[0012] Preferably, the maximum angular force F during small-amplitude rotation in the motion process 转，max =1N;

[0013] Preferably, the maximum angular velocity during normal motion It is 0.78 rad / s.

[0014] A method for identifying and monitoring the tail-wagging state of passenger luggage in a transportation hub includes the following steps:

[0015] Step 1: Install monitoring cameras and loudspeaker alerts in densely populated areas of transportation hubs, and connect the monitoring cameras to the monitoring processor in the control room;

[0016] Step 2: Turn on the monitoring camera, with its shooting direction tangent to the direction of passenger movement, capture video frames, and provide the video data to the monitoring processor;

[0017] Step 3: The monitoring processor calculates the angular force F of the suitcase in the monitoring location using the single-target tracking algorithms MMTrack and findContours respectively. 转,t rate of change of centroid distance ΔT a,t angular velocity of motion

[0018] Step 4: Detect the tail-swing state of the suitcase based on the suitcase tail-swing state recognition model.

[0019] Specifically, step 3 includes:

[0020] Step 3.1: For the video frames captured by the camera, extract the luggage motion information into a four-parameter tuple R using MMTrack. i The angular force F acting on the suitcase is calculated using equations (2) to (8). 转,t ;

[0021] Step 3.2: Extract the outline image of the suitcase using findContours and convert it to grayscale;

[0022] Step 3.3: The i+j order moments of the grayscale image are calculated using equation (16):

[0023] M ij =∑ x ∑ y x i y j G(x,y) (16)

[0024] Among them, M ij Let G(x,y) represent the moment at the origin, i and j represent the order of the moment, x and y represent the pixel coordinates in the image, and G(x,y) represent the pixel gray value at coordinates (x,y).

[0025] Step 3.4: Calculate the zero-order moment M of the grayscale image using equation (16). 00 and first moment M 01 Then passenger (C) x C y ) and suitcase (B x B y The coordinates of the centroid are given by equations (17) and (18):

[0026]

[0027]

[0028] The rate of change of the distance between the center of mass of the suitcase and the passenger is ΔT a,t for:

[0029]

[0030] Where T is the centroid distance between the suitcase and the passenger, ΔT represents the change in centroid distance between the suitcase and the passenger between the i-th frame and the N-th frame, and Δt represents the time interval between the i-th frame and the N-th frame;

[0031] During the tail-swing process of the suitcase, due to the very short time, the rotation angle is recorded as Equation (20), Equation (21), and Equation (22):

[0032]

[0033]

[0034]

[0035] Where Δl is the displacement of the center of mass, Δθ is the turning angle of the suitcase, D is the length of the suitcase handle, and D+C y It is approximately the radius of rotation of the center of mass.

[0036] Specifically, in step 4, the luggage tail-wagging state recognition model is as follows:

[0037]

[0038] Right now:

[0039] 1. The angular force F acting on the suitcase starting from frame t. 转,t The maximum angular force F of small-amplitude rotation during the motion process is greater than that of the maximum rotation force. 转，max ;

[0040] 2. At frame t, the rate of change ΔT of the centroid distance between the suitcase and the passenger. a,t Greater than 0.2 m / s, but less than 0.5 m / s;

[0041] 3. At frame t, the angular velocity of the movement between the suitcase and the passenger. Greater than the maximum angular velocity of normal motion

[0042] The luggage tail-swing state is detected, that is, it is determined whether the luggage tail-swing state recognition model is satisfied.

[0043] Preferably, the maximum angular force F during small-amplitude rotation in the motion process 转，max =1N.

[0044] Preferably, the maximum angular velocity during normal motion It is 0.78 rad / s.

[0045] Furthermore, for the detection of the tail-wagging state of the suitcase, when the above three judgment conditions are met, it can be determined that the suitcase and the passenger have experienced an abnormal tail-wagging state at time t, which may lead to danger. Then, the loudspeaker warning equipment will promptly play and remind pedestrians to pay attention and issue warnings in densely populated areas of transportation hubs until the normal state is restored.

[0046] Otherwise, the suitcase is considered to be in normal motion.

[0047] The beneficial effects of this invention are:

[0048] Unlike other identification methods, this invention identifies the motion characteristics of a suitcase throughout its entire movement and can predict its subsequent motion. Based on the identification results, timely warnings can be issued, effectively preventing potential hazards caused by suitcase tail swinging, which is of great significance for safety management in densely populated transportation hubs. Attached Figure Description

[0049] Figure 1 The steering force experienced by the passenger and luggage in normal straight-line travel.

[0050] Figure 2 The linear force exerted between the passenger and the luggage in a normal straight-line motion.

[0051] Figure 3 The magnitude of the resultant force on the luggage during normal straight-line movement.

[0052] Figure 4 Example: A diagram showing the turning of a suitcase.

[0053] Figure 5 Example: The distance between the center of mass of the suitcase and the passenger.

[0054] Figure 6 Example: Diagram showing the change in distance between two objects during a sudden turn.

[0055] Figure 7 Flowchart of a method for identifying and monitoring the tail-wagging status of passenger luggage in transportation hubs. Detailed Implementation

[0056] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0057] The theoretical research process of the energy-based stability evaluation method for passenger and luggage combined motion systems includes the following steps:

[0058] Step 1: Dynamics Analysis of the Suitcase

[0059] This invention uses the single-target tracking algorithm MMTrack to track the luggage in the captured video, extracting relevant motion information of the suitcase into a four-parameter tuple. Video analysis extracts the luggage's motion trajectory along the frames into a finite four-parameter tuple matrix:

[0060]

[0061] R i This represents the motion data of the suitcase sequence in frame i of the video, where i represents the frame number of the video and j represents the suitcase number. and This represents the x and y coordinates of suitcase j at frame i of the video. and This represents the horizontal and vertical velocities of suitcase j at frame i of the video. The velocity curve shows the magnitude of the resultant velocity of the suitcase, which is... and The norm can be obtained through This can be represented by [the algorithm]. Passenger speed profiles and locations can also be obtained from the passenger online tracking algorithm MMTrack.

[0062] The angle Δθ of the movement of suitcase j from frame i to frame Nj for:

[0063]

[0064] In the i-th frame of the video, the angle θ between the suitcase j and the positive horizontal direction is... j,i satisfy:

[0065]

[0066] Then the angular velocity ω of suitcase j j,Δt , and angular acceleration ε j,Δt for:

[0067]

[0068]

[0069] Δt is the time required for suitcase j to move from frame i to frame N.

[0070] The external torque acting on suitcase j:

[0071] M = J·ε j,Δt =F 转 ·d j (6)

[0072] In the experimental calculations, the suitcase is considered as a cube, and L is denoted as the side length of the cube. Therefore, the moment of inertia of the suitcase is:

[0073]

[0074] m i The mass of the suitcase is uniformly distributed. j This refers to the length of the luggage's rotational torque (usually the length of the luggage plus the handle).

[0075] The external force acting on suitcase j during rotation is:

[0076]

[0077] The horizontal linear acceleration a of suitcase j from frame i to frame N x The vertical linear acceleration and the resultant acceleration 'a' are:

[0078]

[0079]

[0080]

[0081] The magnitude of the external force driving the movement of suitcase j is:

[0082]

[0083] The suitcase moves within the x,y plane during its motion. In the analysis, the suitcase and the passenger move without a height difference in the z-direction. The forces are then transferred to the x,y plane. The external forces acting on the suitcase come from the steering force provided by the passenger, the frictional force from the ground, and the disturbance forces from surrounding passengers. Under normal circumstances, the suitcase is only subject to the external forces exerted by its carrier. Therefore, the external forces acting on the suitcase during normal motion are:

[0084] F 外 =F 转 +F 直 (13)

[0085] Therefore, the external force acting on the suitcase is also the internal force between the suitcase and its carrier. When the suitcase and the passenger move stably together, the system consisting of the suitcase and the carrier is a stable system, and at this time the net external force acting on the suitcase is 0.

[0086] Step Two: Kinematic Analysis of Passengers and Luggage

[0087] It is known that the distance between the center of gravity of the passenger and the large suitcase pulled back under normal conditions is approximately 0.9m. Therefore, the torque of the suitcase is also assumed to be 0.9m, and the mass of the suitcase is assumed to be 15kg. Based on the change in angular acceleration, the magnitude of the steering force at the key frame during the motion can be calculated using equation (9). The steering force is expressed as follows: Figure 1 As shown.

[0088] When the passenger and luggage are in normal motion, |Frotation| ≤ 1N. Therefore, the maximum steering force F is defined as follows: 转，max =1N. In a straight line, the speed of the large suitcase is between 0.8 and 0.9. According to formulas (9)-(12), the forces acting on the suitcase during its linear motion are as follows: Figure 2 As shown.

[0089] It can be seen that when the suitcase is in normal linear motion, the magnitude of the linear force |F_direct| on the suitcase is ≤1N, which is almost zero. That is, during normal motion, the linear force received by the suitcase is almost zero. The suitcase moves at a near-uniform speed as the passenger swings with their steps. The walking speed range of a normal passenger is 1.1m / s to 2m / s, and the running speed is 2m / s to 10m / s. The fastest instantaneous speed of a passenger is 12.42m / s, set by Jamaican sprinter Usain Bolt. We exclude athletes and other special groups. Therefore, the instantaneous acceleration range of a normal passenger is 0 to 5m / s². The acceleration of the suitcase lags behind the passenger's acceleration. When the instantaneous acceleration is too small, it is difficult to detect. Therefore, when the suitcase's acceleration is greater than 1m / s², the acceleration will be considered as follows:2 At this point, the suitcase is considered to be in a state of accelerated motion. Taking the aforementioned suitcase as an example, the critical linear force F acting on the suitcase at this time... 直,max =15N.

[0090] The magnitude of the forces acting on the suitcase during its motion is calculated based on the steering force and the linear force. When the suitcase moves at a constant speed along a straight line, the magnitude of the net force acting on the suitcase is as follows: Figure 3 As shown.

[0091] Depend on Figure 3 and Figure 2 , Figure 1 The comparison shows that the forces acting on the suitcase during its movement are mainly influenced by its steering force. This steering force is primarily generated by the torque provided by the passenger's shoulder joint; further research will investigate the relationship between the passenger's shoulder joint torque and the steering force.

[0092] Step 3: Abnormal State Recognition Model for Suitcases

[0093] An abnormal state refers to any inappropriate action, posture, or event performed by a target in a given scenario. Abnormal states typically possess the following characteristics: unpredictability, environmental relevance, local spatiotemporal nature, non-periodicity, suddenness, short duration, and low frequency. In crowded public places, the abnormal state of a suitcase can disrupt the stability of the crowd. When this disturbance accumulates, it can cause anomalies in the entire public place system, thus threatening public safety. Typical abnormal states of a suitcase include sudden changes in speed, detachment, and changes in direction of movement. This paper will design an identification model for the tail-wagging abnormal state of a suitcase.

[0094] In the abnormal state of a suitcase tail-swing, because the suitcase moves along the ground with the passenger using its wheels, it maintains a certain inertia during movement. When the passenger carrying the suitcase suddenly changes direction, the suitcase will be pulled to turn, but the turning occurs sequentially: the passenger turns first, followed by the suitcase. From a physics perspective, when a multi-connected system is turning, the suitcase at the rear, due to its tendency to maintain its original motion, will experience an outward pulling force, causing it to move away from its carrier. This effect is known as the tail-swing effect of a passenger carrying a suitcase system. Suitcase tail-swing often occurs when passengers make sudden, large turns; therefore, the determination of the suitcase tail-swing effect is as follows:

[0095]

[0096] 1. The angular force F acting on the suitcase starting from frame t. 转,t The maximum angular force of a small-amplitude rotation during the motion can be obtained as F.转，max =1N;

[0097] 2. At frame t, the rate of change ΔT of the centroid distance between the suitcase and the passenger. a,t Greater than 0.2 m / s, but less than 0.5 m / s;

[0098] 3. At frame t, the angular velocity of the movement between the suitcase and the passenger. Greater than the maximum angular velocity of normal motion It is 0.78 rad / s.

[0099] Step 4: Application of the luggage tail-swing state recognition model

[0100] It predicts and monitors the movement of suitcases and issues timely warnings based on the identification results.

[0101] Example 1 and Verification

[0102] This experimental design uses a 28-inch suitcase as an example, and conducts the corresponding experiment inside Shanghai Metro Line 6. Two shooting methods are employed: overhead and side shots. The side shot is positioned 1.2 meters above the ground, centered on the test section, with the passenger moving in a straight line and the camera's shooting direction tangent to the passenger's movement direction. The overhead shot is positioned 5 meters above the ground on the second floor of the metro station, with the camera's image stabilization function activated during shooting to maintain stability.

[0103] To address the luggage tail-wagging phenomenon, 30 frames of motion data were selected to monitor the relevant motion data of the luggage's turning, such as... Figure 4 As shown, (a) is before turning, (b) is during turning, and (c) is after turning.

[0104] This allows us to obtain a graph showing the distance traveled between the suitcase and the passenger, as well as the change in the suitcase's angular velocity. Figure 5 , Figure 6 As shown.

[0105] Depend on Figure 5 As can be seen, the distance between the center of mass of the suitcase and the passenger suddenly increases significantly in frame 15, reaching its peak around frame 19, and then decreases again to return to normal. This is because there is a certain time delay between the suitcase and the passenger's movement. When the suitcase moves normally alongside the passenger, the distance between their centers of mass remains within the range of 50-80mm.

[0106] Depend on Figure 6 It can be seen that around 15 seconds, the angular velocity of the suitcase suddenly increases, reaches its maximum value, and then rapidly decreases back to within the normal range of motion. At frame 15, the angular acceleration experienced by the suitcase is ε. j,Δt= 2.4 rad / s / 0.6 s = 4 rad / s 2 The length of the luggage's rotational torque d j =1.2m. The calculated length of the moment of inertia of the moving obstacle is L = 0.6m. The weight of the suitcase is m. i =18kg. Therefore, the magnitude of the external force acting on the suitcase during rotation is:

[0107]

[0108] In summary, the above verification shows that the angular force on the suitcase starting from frame 15 is greater than the maximum angular force during a small-amplitude rotation, and the rate of change of the center of mass distance ΔT between the suitcase and the passenger is also greater. a,t The angular velocity between the passenger and the luggage is greater than 0.2 m / s but less than 0.5 m / s; the angular velocity is greater than the maximum normal angular velocity of 0.78 rad / s. This indicates that the passenger and luggage assembly is experiencing a steering force exceeding the normal magnitude. Therefore, it can be determined that the passenger and luggage structure is exhibiting an abnormal tail-wagging state, which may lead to danger and should be warned immediately.

[0109] Example 2 and a method for identifying and monitoring the tail-wagging state of passenger luggage in a transportation hub

[0110] like Figure 7 As shown, a method for identifying and monitoring the tail-wagging state of passenger luggage in a transportation hub includes the following steps:

[0111] Step 1: Install monitoring cameras and loudspeaker alerts in densely populated areas of transportation hubs. Connect the monitoring cameras to the monitoring processor in the control room.

[0112] Step 2: Turn on the monitoring camera, with its shooting direction tangent to the direction of passenger movement, collect video frames at a certain frequency, and provide the video data to the monitoring processor.

[0113] Step 3: The monitoring processor calculates the angular force F of the suitcase in the monitoring location using the single target tracking algorithms MMTrack and findContours respectively. 转,t rate of change of centroid distance ΔT a,t angular velocity of motion The specific implementation method is as follows:

[0114] (1) For the video frames captured by the camera, extract the luggage motion information into a four-parameter tuple R using MMTrack. i The angular force F acting on the suitcase is calculated using equations (2) to (8). 转,t ;

[0115] (2) Extract the outline image of the suitcase using findContours and convert it into a grayscale image;

[0116] (3) The i+j moment of the grayscale image can be calculated using equation (16):

[0117] M ij =∑ x ∑ y x i y j G(x,y) (16)

[0118] Among them, M ij Let G(x,y) represent the moment at the origin, i and j represent the order of the moment, x and y represent the pixel coordinates in the image, and G(x,y) represent the pixel gray value at coordinates (x,y).

[0119] (4) In the fields of image processing and computer vision, each planar graphic is composed of pixels, and the centroid coordinates are the weighted average of the coordinates of all the pixels that make up the planar graphic. The image moment is the weighted average of the pixel values ​​of the image. It can be used to find some specific attributes of the image, such as radius, area, centroid, etc. In order to find the centroid of the image, it is usually binarized and then its centroid is found. The zero-order moment M of the grayscale image is calculated by equation (16). 00 and first moment M 01 Then passenger (C) x C y ) and suitcase (B x B y The coordinates of the centroid are given by equations (17) and (18):

[0120]

[0121]

[0122] The rate of change of the distance between the center of mass of the suitcase and the passenger is ΔT a,t for:

[0123]

[0124] Where T is the centroid distance between the suitcase and the passenger, ΔT represents the change in centroid distance between the suitcase and the passenger between the i-th frame and the N-th frame, and Δt represents the time interval between the i-th frame and the N-th frame;

[0125] During the tail-swing process of the suitcase, since the time is very short, the rotation angle can be approximated as Equations (20), (21), and (22):

[0126]

[0127]

[0128]

[0129] Where Δl is the displacement of the center of mass, Δθ is the turning angle of the suitcase, D is the length of the suitcase handle, and D+C y It is approximately the radius of rotation of the center of mass.

[0130] Step 4: Detect the tail-swing state of the suitcase based on the suitcase tail-swing state recognition model. Specifically, determine whether the following conditions are met:

[0131]

[0132] Right now:

[0133] 1. The angular force F acting on the suitcase starting from frame t. 转,t The maximum angular force of a small-amplitude rotation during the motion can be obtained as F. 转，max =1N;

[0134] 2. At frame t, the rate of change ΔT of the centroid distance between the suitcase and the passenger. a,t Greater than 0.2 m / s, but less than 0.5 m / s;

[0135] 3. At frame t, the angular velocity between the suitcase and the passenger is greater than the maximum angular velocity during normal motion. It is 0.78 rad / s.

[0136] For luggage tail-swing detection, if the above three judgment conditions are met, it can be determined that the luggage and passenger have experienced an abnormal tail-swing state at time t, which may lead to danger. Then, the loudspeaker warning equipment will promptly play and remind pedestrians to pay attention and issue warnings in densely populated areas of transportation hubs until the normal state is restored.

[0137] Otherwise, the suitcase is considered to be in normal motion.

[0138] The above description is merely a description of preferred embodiments of this application and is not intended to limit the scope of this application in any way. Any changes or modifications made by those skilled in the art based on the above-disclosed technical content should be considered as equivalent and valid embodiments and fall within the scope of protection of the technical solution of this application.

[0139] References

[0140] [1]Baumgartner T, Mitzel D, Leibe B. Tracking People and Their Objects[J]. IEEE, 2013.

[0141] [2] Dong Daheng. Intelligent analysis of crowd evacuation stability based on video [D]. Tongji University, 2019.

[0142] [3] Mao Xingyun. Introduction to OpenCV Programming [M]. Electronic Industry Press, 2015: 66-68.

Claims

1. A traffic hub passenger luggage case fishtailing state identification and safety monitoring method, characterized in that, The method comprises the steps of: Step 1, setting up monitoring cameras and horn prompting equipment in a passenger flow dense place of a traffic hub, and connecting the monitoring cameras with a monitoring processor in a control room; Step 2, starting the monitoring cameras, the photographing direction of which is tangent to the passenger movement direction, collecting video frames, and providing video data to the monitoring processor; Step 3, the monitoring processor calculates the corner force of the luggage in the monitoring place by single target tracking algorithm MMTrack and findContours respectively , the change rate of the center of mass distance between the luggage and the passenger , the angular velocity of the motion ; Step 4, detecting the tail-swinging state of the luggage based on a luggage tail-swinging state recognition model; The step 3 comprises: Step 3.1, extracting the motion information of the luggage to a four-tuple by MMTrack for the video frame captured by the camera ; calculating the corner force suffered by the luggage by equations (2)-(8) ; Step 3.2, extracting the contour picture of the luggage by findContours and converting it into a gray-scale picture; Step 3.3, calculating the i+j order moment of the gray-scale picture by formula (16): （16） wherein, wherein, m(x, y) represents the origin moment, i and j represent the order of the moment, x and y represent the pixel coordinates in the image, and G(x, y) represents the pixel gray value at the coordinates (x, y). Step 3.

4. Calculate the zeroth moment of the gray scale image by equation (16) and the first moment then the passenger and the suitcase The centroid coordinates are given by equations (17), (18): （17） （18） the rate of change of the centre of mass distance between the luggage and the passenger is: （19） where T is the center of mass distance between the luggage and the passenger, represents the change in the center of mass distance between the luggage and the passenger between the i-th frame and the N-th frame, represents the time interval between the i-th frame and the N-th frame. In the process of the tail-swinging of the luggage, the rotation angle is recorded as formula (20), formula (21) and formula (22) due to the very short time: （20） （21） （22） wherein is the center of mass displacement, is the suitcase corner angle, D is the suitcase handle length, is the approximate center of mass turning radius; In the step 3.1, The four-parameter tuple is: （1） motion data of the luggage sequence in the i-th frame of the video, where i represents the frame number of the video, j represents the luggage number, and represents the horizontal and vertical coordinates of the luggage j in the i-th frame of the video; and represents the horizontal and vertical velocities of the luggage j in the i-th frame of the video; the magnitude of the combined velocity of the luggage is represented by ; the velocity profile and position of the passenger are obtained from the passenger online tracking algorithm MMTrack; Motion angle of the luggage j from the i-th frame to the N-th frame is: （2） j angle of the i-th frame video with the horizontal positive direction satisfies: （3） the angular velocity of the luggage j with the angular acceleration is: （4） （5） time needed for the suitcase j to move from the i-th frame to the N-th frame; The external force distance received by the luggage j: （6） In the experimental calculation process, the luggage is regarded as a cube, L is recorded as the side length of the cube, and the rotational inertia of the luggage is: （7） mass of the luggage case, the mass being evenly distributed; length of the moment of rotation of the luggage case; The turning external force received by the luggage j is: （8）。 2. The method of claim 1, wherein, The step 4, the luggage tail-swinging state recognition model is: （14） That is: cornering force experienced by the luggage from the tth frame , greater than the maximum cornering force for small amplitude turns during motion ; Rate of change of the center of mass distance between the baggage and the passenger at the tth frame greater than 2 m / s, but less than 0.5 m / s; At the t-th frame, the angular velocity of motion between the luggage and the passenger is greater than the maximum angular velocity of normal motion ; Detecting the tail-swinging state of the luggage, that is, judging whether the luggage tail-swinging state recognition model is satisfied.

3. The method of claim 1, wherein, Maximum cornering force of small amplitude rotation during motion = 1 N.

4. The method of claim 1, wherein, Maximum angular velocity of normal movement was 0.78 rad / s.

5. The method of claim 1, wherein, Detecting the tail-swinging state of the luggage, when the three judgment conditions of the luggage tail-swinging state recognition model are satisfied, it can be judged that the luggage and the passenger have the abnormal state of tail-swinging at t time, which may cause danger, then the horn prompting equipment is used to timely play and prompt the pedestrians to pay attention and give early warning in the passenger flow dense place of the traffic hub until the normal state is restored; Otherwise, it is judged that the luggage movement state is normal.

Images