Urban road traffic flow monitoring method based on aerial survey of unmanned aerial vehicle

Abstract

The invention provides an urban road traffic flow monitoring method based on aerial survey of an unmanned aerial vehicle, and the method comprises the steps: S1, shooting a real-time road video of a to-be-monitored road through a camera carried by the unmanned aerial vehicle, and collecting vehicle data in the road; s2, the unmanned aerial vehicle transmits the collected real-time road video of the road to be monitored to a far-end road condition prediction center through an existing communication network; s3, the road condition prediction center processes the collected real-time road video of the road to be monitored to obtain a picture set containing a vehicle; s4, processing the picture set containing the vehicle, and calculating vehicle data; s5, obtaining road parameters; s6, predicting the road condition of the road according to the vehicle data and the road parameters calculated in the step S4; according to the urban road traffic flow monitoring method based on aerial survey of the unmanned aerial vehicle, the vehicles on the road can be counted in real time under the condition of temporarily occupying the existing lane for a long time, so that the vehicles are timely guided to be shunted to reduce traffic congestion.

Description

Technical Field

[0001] The present invention belongs to the technical field of urban road traffic monitoring, and in particular relates to an urban road traffic monitoring method based on unmanned aerial vehicle (UAV) aerial survey. Background Art

[0002] With the rapid advancement of urbanization, the number of motor vehicles in my country has also increased rapidly. Coupled with the surge in the volume of cargo circulation in my country, the accompanying problem of truck overloading has become increasingly serious, which has caused serious damage to infrastructure such as roads and bridges. In addition to increasing a large amount of maintenance costs, there are also a large number of safety hazards. In order to ensure road safety, it is particularly important to repair damaged road surfaces in a timely manner. On the other hand, in order to alleviate the traffic pressure on existing roads, planning new roads is also a common measure.

[0003] However, whether it is repairing existing roads or building new roads, it will cause certain road traffic congestion. If the vehicles traveling on the roads that may be congested can be monitored in real time, the road congestion can be predicted in advance and accurately forecasted.

[0004] The existing urban traffic road monitoring system sets up monitoring points at key places and intersections of urban traffic roads and installs fixed front-end video and audio acquisition equipment to obtain video images and transmit them to the traffic command center through various means (optical fiber, dedicated line, etc.) to store, process and publish information, so that traffic command management personnel can make judgments on traffic violations, traffic jams, traffic accidents and other emergencies, and adjust various system control parameters and command and dispatch strategies accordingly.

[0005] This monitoring system can predict daily traffic conditions in advance, but it cannot timely and accurately respond to abnormal road conditions when predicting emergency traffic conditions. How to solve the problem of monitoring and forecasting temporary road conditions is still a technical problem that needs to be solved urgently. Summary of the invention

[0006] The purpose of the present invention is to provide a method for monitoring urban road traffic based on drone aerial survey, so as to solve the problems of real-time traffic condition monitoring and road condition prediction of road sections with temporary situations.

[0007] To achieve the above-mentioned purpose, the present invention provides a method for monitoring urban road traffic flow based on drone aerial survey, comprising the following steps:

[0008] Step S1: Use the camera carried by the drone to shoot real-time road video of the road to be monitored and collect vehicle data on the road;

[0009] Step S2: The drone uses the existing communication network to transmit the collected real-time road video of the road to be monitored to a remote road condition prediction center;

[0010] Step S3: The road condition prediction center processes the collected real-time road video of the road to be monitored to obtain a picture set containing vehicles;

[0011] Step S4, processing the picture set containing the vehicle and calculating the vehicle data;

[0012] Step S5, obtaining road parameters;

[0013] Step S6: predicting the road condition based on the vehicle data and road parameters calculated in step S4.

[0014] Furthermore, the specific process of step S1, using the camera carried by the drone to shoot real-time road video of the road to be monitored and collect vehicle data on the road; step S2, the drone uses the existing communication network to transmit the collected real-time road video of the road to be monitored to the remote road condition prediction center is: the road condition prediction center sends road monitoring parameters to the drone through the existing communication network according to the needs of road condition monitoring; the drone completes the real-time road video collection and transmission according to the received flight and monitoring instructions, and then returns.

[0015] Furthermore, in step S3, the road condition prediction center processes the collected real-time road video of the road to be monitored to obtain a picture set containing vehicles in the following specific process:

[0016] Step S301: first, use Python to convert the collected real-time road video of the road to be monitored into a picture;

[0017] Step S302: filter the pictures obtained in step S301, delete the pictures that do not contain vehicles, and save the remaining pictures in order as a picture set.

[0018] Furthermore, the specific process of step S302, screening the pictures obtained in step S301, is: comparing the infrared images and visible light images synchronously taken by the drone frame by frame, if there is an infrared target in the infrared image, retaining the corresponding infrared image and visible light image, otherwise deleting the infrared image and visible light image of the corresponding frame; and numbering the retained images to form an infrared image set M1 and a visible light image set M2;

[0019] The corresponding images in the infrared image set M1 and the visible light image set M2 are fused to obtain a fused image set M.

[0020] Furthermore, the specific process of step S4, processing the image set containing the vehicle and calculating the vehicle data is:

[0021] The images of the picture set M containing vehicles are input into the GS-SSD vehicle detection network model, and the GS-SSD vehicle detection network is used to perform target detection on the obtained images, and finally the number of vehicles counted is output.

[0022] Furthermore, the step S5 obtains road parameters, including road occupation parameters, road traffic parameters, and weather parameters.

[0023] Furthermore, in step S6, the road condition is predicted based on the vehicle data and road parameters calculated in step S4, and the road congestion time is predicted using the following formula:

[0024]

[0025] Among them, T is the congestion time of the road section, t is the time when the vehicle passes at a safe speed v; λ and γ are adjustment parameters, and the specific values ​​are adjusted according to the actual road section information. i is the traffic flow slope of the i-th segment, k is the average traffic flow slope of the D segment, D is the total number of time periods; when the traffic flow slope K of the i-th section i If it is smaller than the traffic flow slope k of the road section, the second term in the formula is negative, and the congestion formation time is relatively reduced. On the contrary, the traffic flow slope K of the i-th section i If it is smaller than the traffic flow slope k of the road section, the congestion formation time will increase relatively; i is the proportion of traffic flow that continuously exceeds the reference value during the i-th period, p is the average value of the proportion of traffic flow that continuously exceeds the reference value during the D-th period, P i The relationship between β and p is the same as that between Ki and k; n represents the bias coefficient of the nth road parameter, γ n Indicates the influence of the nth road parameter;

[0026] Among them, k i =M i -M i-1 , where M i represents the vehicle flow in the i-th period; M i-1 represents the vehicle flow in the i-1th period;

[0027] Among them, N j It represents the vehicle flow in the jth period of time; N0 represents the reference value of vehicle flow, and D represents the total amount in the time period.

[0028] The advantages of the present invention are as follows: the present invention provides a method for monitoring the traffic flow of urban roads based on drone aerial survey, which can take photos / videos along the same path at a fixed time and fixed point to count the traffic flow in each time period. The traffic flow statistics of each time are recorded to form a road traffic monitoring database, and the traffic flow in a certain time period in the future can be predicted based on the monitoring data. In particular, in emergencies, such as road repairs, new roads, etc., where existing lanes are temporarily occupied for a long time, the vehicles on the road can be counted in real time, and the road congestion can be reported in time, so as to guide the vehicles to divert in time and minimize traffic congestion.

[0029] The present invention is described in detail below with reference to the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 This is a flow chart of a method for monitoring urban road traffic based on UAV aerial survey. DETAILED DESCRIPTION

[0031] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined purpose, the specific implementation methods, structural features and effects of the present invention are described in detail below in conjunction with the accompanying drawings and embodiments.

[0032] 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.

[0033] In the description of the present invention, it is necessary to understand that the terms "center", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "aligned", "overlap", "bottom", "inside", "outside" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present invention.

[0034] The terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features; in the description of the present invention, unless otherwise specified, "plurality" means two or more.

[0035] Example 1

[0036] This embodiment provides a Figure 1 The method for monitoring urban road traffic based on drone aerial survey shown in the figure includes the following steps:

[0037] Step S1: Use the camera carried by the drone to shoot real-time road video of the road to be monitored and collect vehicle data on the road;

[0038] Step S2: The drone uses the existing communication network to transmit the collected real-time road video of the road to be monitored to a remote road condition prediction center;

[0039] Step S3: The road condition prediction center processes the collected real-time road video of the road to be monitored to obtain a picture set containing vehicles;

[0040] Step S4, processing the picture set containing the vehicle and calculating the vehicle data;

[0041] Step S5, obtaining road parameters;

[0042] Step S6: predicting the road condition based on the vehicle data and road parameters calculated in step S4.

[0043] Furthermore, the specific process of step S1, using the camera carried by the drone to shoot real-time road video of the road to be monitored and collect vehicle data on the road; step S2, the drone uses the existing communication network to transmit the collected real-time road video of the road to be monitored to the remote road condition prediction center is: the road condition prediction center sends road monitoring parameters and instructions to the drone through the existing communication network according to the needs of road condition monitoring; the drone completes the real-time road video collection and transmission according to the received flight, monitoring instructions and parameters, and then returns.

[0044] Furthermore, in step S3, the road condition prediction center processes the collected real-time road video of the road to be monitored to obtain a picture set containing vehicles in the following specific process:

[0045] Step S301: first, use Python software to convert the collected real-time road video of the road to be monitored into a picture;

[0046] Step S302: filter the pictures obtained in step S301, delete the pictures that do not contain vehicles, and save the remaining pictures in order as a picture set.

[0047] Furthermore, the specific process of step S302, screening the pictures obtained in step S301, is: comparing the infrared images and visible light images synchronously taken by the drone frame by frame; if there is an infrared target in the infrared image, the corresponding infrared image and visible light image are retained; otherwise, the infrared image and visible light image of the corresponding frame are deleted; and the retained images are numbered to form an infrared image set M1 and a visible light image set M2.

[0048] The corresponding images in the infrared image set M1 and the visible light image set M2 are fused to obtain a fused image set M.

[0049] The fusion method of the infrared image set M1 and the visible light image set M2 can adopt the existing image method, which will not be described in detail in this embodiment.

[0050] Furthermore, the specific process of step S4, processing the image set containing the vehicle and calculating the vehicle data is:

[0051] The images of the picture set M containing vehicles are input into the GS-SSD vehicle detection network model, and the GS-SSD vehicle detection network is used to perform target detection on the obtained images, and finally the number of vehicles counted is output.

[0052] The GS-SSD vehicle detection network model is built based on the SSD vehicle detection network model. The patent application with application number 2020108624795 discloses its specific scheme for detecting vehicles, which will not be described in detail in this example.

[0053] Furthermore, the step S5 obtains road parameters, including road occupation parameters, road traffic parameters, and weather parameters.

[0054] Furthermore, in step S6, the road condition is predicted based on the vehicle data and road parameters calculated in step S4, and the road congestion time is predicted using the following formula:

[0055]

[0056] Among them, T is the congestion time of the road section, t is the time that the vehicle passes at a safe speed v; λ and γ are adjustment parameters, λ = 1-γ, and the specific value is adjusted according to the actual road section information, K i is the traffic flow slope of the i-th segment, k is the average traffic flow slope of the D segment, D is the total number of time periods; when the traffic flow slope K of the i-th section i If it is smaller than the traffic flow slope k of the road section, the second term in the formula is negative, and the congestion formation time is relatively reduced. On the contrary, the traffic flow slope K of the i-th section i If it is smaller than the traffic flow slope k of the road section, the congestion formation time will increase relatively; iis the proportion of traffic flow that continuously exceeds the reference value during the i-th period, p is the average value of the proportion of traffic flow that continuously exceeds the reference value during the D-th period, P i The relationship between β and p is the same as that between Ki and k; n Represents the bias coefficient of the nth road parameter, with values ​​of (0, 1, 2, 3, 4), γ n It indicates that the influence value of the nth road parameter is (4, 3, 2, 1, 0).

[0057] Among them, k i =M i -M i-1 , where M i represents the vehicle flow in the i-th period; M i-1 represents the vehicle flow in the i-1th period;

[0058] Among them, N j It represents the vehicle flow in the jth period of time; N0 represents the reference value of vehicle flow, and D represents the total amount in the time period.

[0059] In summary, the present invention provides a method for monitoring urban road traffic based on drone aerial survey. The method for monitoring urban road traffic based on drone aerial survey takes photos / videos along the same path at a fixed time and fixed point to count the traffic flow in each time period. Each traffic flow statistics is recorded to form a road traffic monitoring database. The traffic flow in a certain period in the future can be predicted based on the monitoring data. Especially in emergencies, such as road repairs, new roads, etc., where existing lanes are temporarily occupied for a long time, the vehicles on the road can be counted in real time, and the road congestion can be reported in time, so as to guide vehicles to divert in time and minimize traffic congestion.

[0060] Example 2

[0061] The urban road traffic flow monitoring method based on UAV aerial survey is used to monitor the traffic flow on a road that is undergoing road maintenance and the time it takes to pass through the road section during rush hour.

[0062] The road condition prediction center issues road traffic monitoring instructions and traffic parameters (monitoring range: 1 km, monitoring times: 3 times, monitoring interval 30 minutes); the drone carries a camera to shoot real-time road video of the road to be monitored according to the predetermined route, and uses the existing communication network to transmit the collected real-time road video of the road to be monitored to the remote road condition prediction center; the road condition prediction center processes the collected real-time road video of the road to be monitored, obtains a picture set containing vehicles, processes the picture set containing vehicles, and calculates vehicle data. The calculation results show that the number of vehicles calculated in the first collection is 158, the number of vehicles calculated in the second collection is 132, and the number of vehicles calculated in the third collection is 172.

[0063] The road parameters obtained by the Road Condition Prediction Center are currently as follows: the number of occupied lanes on the road is 2, the bias coefficient is 2, and the impact degree is 3; the total number of lanes on the road is 4, the bias coefficient is 1, and the impact degree is 2; the weather is cloudy, the bias coefficient is 2, and the impact degree is 2.

[0064] During rush hour, set λ to 0.6 and γ to 0.4.

[0065] According to the vehicle data and road parameters calculated in the steps, the road conditions are predicted, and the road congestion time is predicted using the following formula:

[0066]

[0067] After calculation, the road section occupies 2 lanes for road construction, and the passing time during rush hour is 23.4 minutes, which is more than 18 minutes longer than the normal passing time of 5 minutes.

[0068] The above contents are further detailed descriptions of the present invention in combination with specific preferred embodiments, and it cannot be determined that the specific implementation of the present invention is limited to these descriptions. For ordinary technicians in the technical field to which the present invention belongs, several simple deductions or substitutions can be made without departing from the concept of the present invention, which should be regarded as falling within the protection scope of the present invention.

Claims

1. A method for monitoring urban road traffic based on drone aerial survey, characterized in that: The steps include: Step S1: Use the camera carried by the drone to shoot real-time road video of the road to be monitored and collect vehicle data on the road; Step S2: The drone uses the existing communication network to transmit the collected real-time road video of the road to be monitored to a remote road condition prediction center; Step S3: The road condition prediction center processes the collected real-time road video of the road to be monitored to obtain a picture set containing vehicles; Step S4, processing the picture set containing the vehicle and calculating the vehicle data; Step S5, obtaining road parameters; Step S6: predicting the road condition based on the vehicle data and road parameters calculated in step S4.

2. The method for monitoring urban road traffic flow based on drone aerial survey as claimed in claim 1, characterized in that: The specific process of step S1, using the camera carried by the drone to shoot real-time road video of the road to be monitored and collect vehicle data on the road; step S2, the drone uses the existing communication network to transmit the collected real-time road video of the road to be monitored to the remote road condition prediction center is: the road condition prediction center sends road monitoring parameters to the drone through the existing communication network according to the needs of road condition monitoring; the drone completes the real-time road video collection and transmission according to the received flight and monitoring instructions, and then returns.

3. The method for monitoring urban road traffic flow based on drone aerial survey as claimed in claim 1, characterized in that: The specific process of the step S3, the road condition prediction center processes the collected real-time road video of the road to be monitored to obtain a picture set containing vehicles is as follows: Step S301: first, use Python to convert the collected real-time road video of the road to be monitored into a picture; Step S302: filter the pictures obtained in step S301, delete the pictures that do not contain vehicles, and save the remaining pictures in order as a picture set.

4. The method for monitoring urban road traffic flow based on drone aerial survey as claimed in claim 3, characterized in that: The specific process of step S302, screening the pictures obtained in step S301, is: comparing the infrared images and visible light images synchronously taken by the drone frame by frame, if there is an infrared target in the infrared image, retaining the corresponding infrared image and visible light image, otherwise deleting the infrared image and visible light image of the corresponding frame; and numbering the retained images to form an infrared image set M1 and a visible light image set M2; The corresponding images in the infrared image set M1 and the visible light image set M2 are fused to obtain a fused image set M.

5. The method for monitoring urban road traffic flow based on drone aerial survey as claimed in claim 1, characterized in that: The specific process of step S4, processing the picture set containing the vehicle and calculating the vehicle data is: The images of the picture set M containing vehicles are input into the GS-SSD vehicle detection network model, and the GS-SSD vehicle detection network is used to perform target detection on the obtained images, and finally the number of vehicles counted is output.

6. The method for monitoring urban road traffic flow based on drone aerial survey as claimed in claim 1, characterized in that: The step S5 is to obtain road parameters, including road occupation parameters, road traffic parameters, and weather parameters.

7. The method for monitoring urban road traffic flow based on drone aerial survey as claimed in claim 1, characterized in that: In step S6, the road condition is predicted based on the vehicle data and road parameters calculated in step S4, and the road congestion time is predicted using the following formula: Among them, T is the congestion time of the road section, t is the time when the vehicle passes at a safe speed v; λ and γ are adjustment parameters, and the specific values ​​are adjusted according to the actual road section information. i is the traffic flow slope of the i-th segment, k is the average traffic flow slope of the D segment, D is the total number of time periods; when the traffic flow slope K of the i-th section i If it is smaller than the traffic flow slope k of the road section, the second term in the formula is negative, and the congestion formation time is relatively reduced. On the contrary, the traffic flow slope K of the i-th section i If it is smaller than the traffic flow slope k of the road section, the congestion formation time will increase relatively; i is the proportion of traffic flow that continuously exceeds the reference value during the i-th period, p is the average value of the proportion of traffic flow that continuously exceeds the reference value during the D-th period, P i The relationship between β and p is the same as that between Ki and k; n represents the bias coefficient of the nth road parameter, γ n Indicates the influence of the nth road parameter; Among them, k i =M i -M i-1 , where M i represents the vehicle flow in the i-th period; M i-1 represents the vehicle flow in the i-1th period; Among them, N j It represents the vehicle flow in the jth period of time; N0 represents the reference value of vehicle flow, and D represents the total amount in the time period.

Images