Traffic section dynamic resource allocation system

Abstract

The present invention relates to a dynamic resource allocation system for traffic sections, comprising: a quantity prediction mechanism, which is used to take the arrived time partition as a target time partition when the time partition belonging to the peak time period of traffic at the intersection arrives every day, and intelligently predict the number of reference personnel corresponding to the target time partition based on the number of reference personnel corresponding to multiple time partitions of a set number before the target time partition, the coverage area of ​​the intersection, and the total number of traffic sections intersecting at the intersection; a strategy customization device, which is used to dispatch different numbers of monitoring resources and management resources to the intersection based on the number of reference personnel corresponding to the target time partition. Through the present invention, the number of pedestrians in the current time partition of the traffic section intersection can be intelligently predicted, and then the dynamic allocation of monitoring resources and management resources for the intersection can be realized based on the prediction results, thereby improving the management level of the traffic section.

Description

Technical Field

[0001] The invention relates to the field of traffic section management, and in particular to a traffic section dynamic resource allocation system. Background Art

[0002] The intersections of traffic sections are the gathering and dispersing points for pedestrians and vehicles, and are more likely to gather a large number of pedestrians and vehicles than other areas in the city. They are also key areas that are difficult for urban management to manage. For example, during the peak hours of each day, the number of pedestrians and vehicles gathered at the intersections in the city is greater. If the traffic is not smooth, the monitoring is ineffective, or traffic accidents occur at these intersections, it will inevitably cause large-scale traffic jams. The more intersections with traffic jams, the lower the traffic flow of the entire city, which makes urban management during peak hours obviously difficult.

[0003] Among the invention patents that have been published, for example:

[0004] The invention with application number 2021114621290 (publication number CN114093179A) relates to a vehicle dispatching method, cloud server, equipment and storage medium at an intersection. The method confirms the first reference direction, the second reference direction, the third reference direction and the fourth reference direction of the intersection, counts the number of vehicles in the target area in the first reference direction, the second reference direction, the third reference direction and the fourth reference direction, and takes the maximum number of vehicles as the target number of vehicles, respectively calculates the total travel time of the target number of vehicles in the lanes corresponding to the two reference directions, sends the total travel time to the traffic light, and controls the traffic light to dispatch each vehicle according to the preset release rules. Since the travel time is calculated based on the maximum number of vehicles, the maximum number of vehicles can reflect the traffic flow of each lane at the intersection within a certain period of time. Using this as a standard to dynamically allocate the traffic time and prohibition time of the traffic light in each lane can well solve the traffic congestion and improve traffic efficiency.

[0005] The invention with application number 2021103582184 (publication number CN113096398A) discloses a comprehensive traffic data mining method of multi-source data fusion, which belongs to the field of data processing technology. The specific steps of the data mining method are as follows: (1) traffic information collection; (2) data preprocessing; (3) feature extraction; (4) construction of initial prediction model; (5) decision-level data fusion; (6) construction of fused traffic prediction model; (7) model evaluation; (8) model application; Compared with the traditional single data acquisition method, the present invention adopts multi-source data sensors for comprehensive traffic data analysis and processing, and constructs a fused traffic prediction model to predict congestion at traffic intersections, which is beneficial to improve the prediction accuracy of traffic congestion, and then facilitates the traffic department to conduct traffic diversion through the traffic information release platform.

[0006] The invention with application number 2020111952761 (publication number CN112365723A) discloses an adaptive calculation method for real-time numerical values ​​of segmented intervals of a traffic light intersection, which relates to the field of intelligent transportation technology, and includes the following steps: detecting the vehicle capacity of a lane traveling toward an intersection in a first direction and the vehicle capacity of a lane traveling toward an intersection in a second direction, comparing L1 with L2, and adaptively jumping the time of the traffic light according to the comparison result and in combination with a preset green light duration and a preset waiting duration, so that the vehicle traffic efficiency is higher; detecting the number of first stranded vehicles entering the intersection along the first direction and being stranded, so as to avoid vehicles from both directions being stranded at the intersection, causing vehicle backlog and serious traffic congestion at the intersection, so as to improve the vehicle traffic efficiency; detecting the number of first stranded pedestrians entering the zebra crossing along the first direction and being stranded, so as to avoid pedestrians being stranded at the zebra crossing when the red light is changed, and vehicles coming from the opposite direction causing traffic accidents, thereby protecting the safety of pedestrians.

[0007] In the prior art, in order to avoid the above dilemma, city managers usually choose to allocate excessive monitoring resources and management resources to each intersection. For example, during peak hours, cameras that were originally not facing the intersection are changed to face the intersection, and a large number of remote managers are allocated to the monitoring server and the monitoring display to perform remote management based on network communication. However, due to the randomness of the number of pedestrians at the intersection, the allocation of the above-mentioned sufficient monitoring and management resources is wasted. On the contrary, if sufficient monitoring and management resources are not allocated, unfavorable monitoring and management scenarios will be formed in unpredictable extreme situations. Summary of the invention

[0008] In order to solve the above problems, the present invention provides a dynamic resource allocation system for traffic sections, which can intelligently predict the number of pedestrians in the current time zone according to the coverage area of ​​the intersection of the traffic section, the number of intersecting traffic sections and the number of pedestrians in historical time, and then realize dynamic allocation of monitoring resources and management resources for the intersection based on the prediction results, thereby avoiding traffic jams at peak times and improving the utilization of limited urban resources.

[0009] According to one aspect of the present invention, a traffic section dynamic resource allocation system is provided, the system comprising:

[0010] An interval acquisition mechanism is arranged above a certain intersection of a traffic section, and is used to perform continuous image acquisition actions at preset time intervals on the environment of the intersection in a bird's-eye view mode, so as to obtain a plurality of interval acquisition frames in each time partition;

[0011] A signal analysis mechanism is connected to the interval acquisition mechanism and is used to perform the following signal analysis processing for each time partition: detecting the number of human imaging areas in each interval acquisition frame belonging to the time partition, and performing deduplication processing on each human imaging area in multiple interval acquisition frames within the time partition to obtain the total number of the remaining multiple human imaging areas as the reference number of people corresponding to the time partition;

[0012] A quantity prediction mechanism is used to, when a time partition belonging to a peak time period of traffic at an intersection is reached on each day, use the arrived time partition as a target time partition, use a plurality of reference numbers of personnel corresponding to a plurality of time partitions of a set number before the target time partition, the coverage area of ​​the intersection, and the total number of traffic sections intersecting at the intersection as a plurality of input data of an artificial intelligence prediction body, so as to run the artificial intelligence prediction body and obtain the number of personnel output by the artificial intelligence prediction body as the reference number of personnel corresponding to the target time partition;

[0013] A learning execution device, connected to the quantity prediction mechanism, for sending the deep neural network after completing a fixed number of learning operations as an artificial intelligence prediction body to the quantity prediction mechanism for use;

[0014] A strategy customization device, connected to the quantity prediction mechanism, for dispatching different quantities of monitoring resources and management resources to the intersection based on the reference number of personnel corresponding to the target time partition;

[0015] Wherein, in the quantity prediction mechanism, the value of the set number is positively correlated with the value of the preset time interval;

[0016] The method of sending the deep neural network after completing a fixed number of learning operations as an artificial intelligence prediction body to the quantity prediction mechanism for use includes: the smaller the coverage area of ​​the intersection, the smaller the value of the fixed number;

[0017] Wherein, dispatching different numbers of monitoring resources and management resources to the intersection based on the number of reference personnel corresponding to the target time partition further includes: the more the number of reference personnel corresponding to the target time partition, the more the number of remote management personnel allocated to the intersection;

[0018] Wherein, dispatching different numbers of monitoring resources and management resources to the intersection based on the number of reference personnel corresponding to the target time partition includes: the more the number of reference personnel corresponding to the target time partition, the more monitoring cameras allocated to the intersection;

[0019] Among them, the more reference personnel the target time partition corresponds to, the more surveillance cameras are allocated to the intersection, including: one surveillance camera can be shared by multiple intersections, and when the monitoring direction of a surveillance camera is the intersection, it is determined that the surveillance camera is allocated to the intersection. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The embodiments of the present invention will be described below with reference to the accompanying drawings, wherein:

[0021] Figure 1 1 is an internal structure diagram of a traffic section dynamic resource allocation system according to Embodiment 1 of the present invention.

[0022] Figure 2 1 is an internal structure diagram of a traffic section dynamic resource allocation system according to Embodiment 2 of the present invention.

[0023] Figure 3 1 is an internal structure diagram of a traffic section dynamic resource allocation system according to Embodiment 3 of the present invention. DETAILED DESCRIPTION

[0024] The implementation scheme of the traffic section dynamic resource allocation system of the present invention will be described in detail below with reference to the accompanying drawings.

[0025] Example 1

[0026] Figure 1 The internal structure diagram of the traffic section dynamic resource allocation system according to Embodiment 1 of the present invention is shown, and the system includes:

[0027] An interval acquisition mechanism is arranged above a certain intersection of a traffic section, and is used to perform continuous image acquisition actions at preset time intervals on the environment of the intersection in a bird's-eye view mode, so as to obtain a plurality of interval acquisition frames in each time partition;

[0028] A signal analysis mechanism is connected to the interval acquisition mechanism and is used to perform the following signal analysis processing for each time partition: detecting the number of human imaging areas in each interval acquisition frame belonging to the time partition, and performing deduplication processing on each human imaging area in multiple interval acquisition frames within the time partition to obtain the total number of the remaining multiple human imaging areas as the reference number of people corresponding to the time partition;

[0029] A quantity prediction mechanism is used to, when a time partition belonging to a peak time period of traffic at an intersection is reached on each day, use the arrived time partition as a target time partition, use a plurality of reference numbers of personnel corresponding to a plurality of time partitions of a set number before the target time partition, the coverage area of ​​the intersection, and the total number of traffic sections intersecting at the intersection as a plurality of input data of an artificial intelligence prediction body, so as to run the artificial intelligence prediction body and obtain the number of personnel output by the artificial intelligence prediction body as the reference number of personnel corresponding to the target time partition;

[0030] A learning execution device, connected to the quantity prediction mechanism, for sending the deep neural network after completing a fixed number of learning operations as an artificial intelligence prediction body to the quantity prediction mechanism for use;

[0031] A strategy customization device, connected to the quantity prediction mechanism, for dispatching different quantities of monitoring resources and management resources to the intersection based on the reference number of personnel corresponding to the target time partition;

[0032] Wherein, in the quantity prediction mechanism, the value of the set number is positively correlated with the value of the preset time interval;

[0033] For example, the value of the set number is positively associated with the value of the preset time interval, including: when the preset time interval is 30 seconds, the value of the set number is 20, when the preset time interval is 60 seconds, the value of the set number is 25, and when the preset time interval is 120 seconds, the value of the set number is 30;

[0034] The method of sending the deep neural network after completing a fixed number of learning operations as an artificial intelligence prediction body to the quantity prediction mechanism for use includes: the smaller the coverage area of ​​the intersection, the smaller the value of the fixed number;

[0035] Wherein, dispatching different numbers of monitoring resources and management resources to the intersection based on the number of reference personnel corresponding to the target time partition further includes: the more the number of reference personnel corresponding to the target time partition, the more the number of remote management personnel allocated to the intersection;

[0036] Wherein, dispatching different numbers of monitoring resources and management resources to the intersection based on the number of reference personnel corresponding to the target time partition includes: the more the number of reference personnel corresponding to the target time partition, the more monitoring cameras allocated to the intersection;

[0037] Among them, the more reference personnel the target time partition corresponds to, the more surveillance cameras are allocated to the intersection, including: one surveillance camera can be shared by multiple intersections, and when the monitoring direction of a surveillance camera is the intersection, it is determined that the surveillance camera is allocated to the intersection.

[0038] Example 2

[0039] Figure 2 is an internal structure diagram of a traffic section dynamic resource allocation system according to Embodiment 2 of the present invention, and Figure 1 different, Figure 2 The dynamic resource allocation system for traffic segments in the system can include the following components:

[0040] An interval acquisition mechanism is arranged above a certain intersection of a traffic section, and is used to perform continuous image acquisition actions at preset time intervals on the environment of the intersection in a bird's-eye view mode, so as to obtain a plurality of interval acquisition frames in each time partition;

[0041] A signal analysis mechanism is connected to the interval acquisition mechanism and is used to perform the following signal analysis processing for each time partition: detecting the number of human imaging areas in each interval acquisition frame belonging to the time partition, and performing deduplication processing on each human imaging area in multiple interval acquisition frames within the time partition to obtain the total number of the remaining multiple human imaging areas as the reference number of people corresponding to the time partition;

[0042] A quantity prediction mechanism is used to, when a time partition belonging to a peak time period of traffic at an intersection is reached on each day, use the arrived time partition as a target time partition, use a plurality of reference numbers of personnel corresponding to a plurality of time partitions of a set number before the target time partition, the coverage area of ​​the intersection, and the total number of traffic sections intersecting at the intersection as a plurality of input data of an artificial intelligence prediction body, so as to run the artificial intelligence prediction body and obtain the number of personnel output by the artificial intelligence prediction body as the reference number of personnel corresponding to the target time partition;

[0043] A learning execution device, connected to the quantity prediction mechanism, for sending the deep neural network after completing a fixed number of learning operations as an artificial intelligence prediction body to the quantity prediction mechanism for use;

[0044] A strategy customization device, connected to the quantity prediction mechanism, for dispatching different quantities of monitoring resources and management resources to the intersection based on the reference number of personnel corresponding to the target time partition;

[0045] Wherein, in the quantity prediction mechanism, the value of the set number is positively correlated with the value of the preset time interval;

[0046] The method of sending the deep neural network after completing a fixed number of learning operations as an artificial intelligence prediction body to the quantity prediction mechanism for use includes: the smaller the coverage area of ​​the intersection, the smaller the value of the fixed number;

[0047] A time detection device, connected to the quantity prediction mechanism, is used to determine the arrival of a time partition belonging to a peak time period for traffic at the intersection on a certain day when the current time is equal to the starting time of the time partition belonging to a peak time period for traffic at the intersection on a certain day;

[0048] Among them, the peak traffic hours at different intersections in each city are also different.

[0049] Example 3

[0050] Figure 3 is an internal structure diagram of a traffic section dynamic resource allocation system according to Embodiment 3 of the present invention, and Figure 1 different, Figure 3 The dynamic resource allocation system for traffic segments in the system can include the following components:

[0051] An interval acquisition mechanism is arranged above a certain intersection of a traffic section, and is used to perform continuous image acquisition actions at preset time intervals on the environment of the intersection in a bird's-eye view mode, so as to obtain a plurality of interval acquisition frames in each time partition;

[0052] A signal analysis mechanism is connected to the interval acquisition mechanism and is used to perform the following signal analysis processing for each time partition: detecting the number of human imaging areas in each interval acquisition frame belonging to the time partition, and performing deduplication processing on each human imaging area in multiple interval acquisition frames within the time partition to obtain the total number of the remaining multiple human imaging areas as the reference number of people corresponding to the time partition;

[0053] A quantity prediction mechanism is used to, when a time partition belonging to a peak time period of traffic at an intersection is reached on each day, use the arrived time partition as a target time partition, use a plurality of reference numbers of personnel corresponding to a plurality of time partitions of a set number before the target time partition, the coverage area of ​​the intersection, and the total number of traffic sections intersecting at the intersection as a plurality of input data of an artificial intelligence prediction body, so as to run the artificial intelligence prediction body and obtain the number of personnel output by the artificial intelligence prediction body as the reference number of personnel corresponding to the target time partition;

[0054] A learning execution device, connected to the quantity prediction mechanism, for sending the deep neural network after completing a fixed number of learning operations as an artificial intelligence prediction body to the quantity prediction mechanism for use;

[0055] A strategy customization device, connected to the quantity prediction mechanism, for dispatching different quantities of monitoring resources and management resources to the intersection based on the reference number of personnel corresponding to the target time partition;

[0056] Wherein, in the quantity prediction mechanism, the value of the set number is positively correlated with the value of the preset time interval;

[0057] The method of sending the deep neural network after completing a fixed number of learning operations as an artificial intelligence prediction body to the quantity prediction mechanism for use includes: the smaller the coverage area of ​​the intersection, the smaller the value of the fixed number;

[0058] A dynamic storage chip is connected to the learning execution device and is used to store various operating parameters of the artificial intelligence prediction body.

[0059] Next, the specific structure of the traffic section dynamic resource allocation system of the present invention will be further described.

[0060] In the traffic section dynamic resource allocation system according to any embodiment of the present invention:

[0061] The method of using a bird's-eye view mode to perform continuous image acquisition actions at preset time intervals on the environment where the intersection is located to obtain a plurality of interval acquisition frames in each time partition includes: adjusting the bird's-eye view angle of the interval acquisition mechanism so that the bird's-eye view of the interval acquisition mechanism includes the intersection;

[0062] Among them, using the overhead mode to perform continuous image acquisition actions at preset time intervals on the environment where the intersection is located to obtain multiple interval acquisition frames in each time partition also includes: adjusting the overhead viewing angle size of the interval acquisition mechanism so that the overhead viewing field of the interval acquisition mechanism takes the intersection as the viewing center;

[0063] Among them, adjusting the overhead viewing angle of the interval acquisition mechanism so that the overhead viewing field of the interval acquisition mechanism takes the intersection as the center of the field of view includes: the centroid position of the imaging area of ​​the intersection in each interval acquisition frame is at the center position of each interval acquisition frame.

[0064] And in the traffic section dynamic resource allocation system according to any embodiment of the present invention:

[0065] Deduplication processing is performed on each human imaging region in the multiple interval acquisition frames within the time partition to obtain the total number of the remaining multiple human imaging regions and use it as the number of reference persons corresponding to the time partition, including: performing deduplication processing on the multiple human imaging regions with matching region edge shapes as the same human imaging region to retain only one human imaging region among the multiple human imaging regions for participating in the subsequent accumulation of the total number of the remaining multiple human imaging regions;

[0066] Wherein, performing deduplication processing on each human imaging region in the multiple interval acquisition frames in the time partition to obtain the total number of the remaining multiple human imaging regions and use it as the number of reference persons corresponding to the time partition further comprises: performing identification of the human imaging region in each interval acquisition frame in the time partition based on the reference outline of the human body;

[0067] Among them, deduplication processing is performed on each human imaging area in the multiple interval acquisition frames within the time partition to obtain the total number of the remaining multiple human imaging areas and use it as the number of reference personnel corresponding to the time partition, which also includes: each human imaging area in the multiple interval acquisition frames within the time partition comes from a certain interval acquisition frame among the multiple interval acquisition frames.

[0068] In addition, in the dynamic resource allocation system for traffic sections, a continuous image acquisition action at a preset time interval is performed on the environment of the intersection in a bird's-eye view mode to obtain multiple interval acquisition frames in each time partition, which also includes: the preset time interval is less than or equal to a set time length threshold, the multiple interval acquisition frames in each time partition are interval acquisition frames whose acquisition timestamps fall within the time partition, and the multiple interval acquisition frames in each time partition have the same resolution.

[0069] According to the embodiments of the present invention, the present invention has at least the following two important inventive concepts:

[0070] Inventive concept A: The number of people flow in the past time zones of the intersection of the traffic section, the intersection coverage area and the total number of traffic sections intersecting at the intersection are used to intelligently predict the number of people flow in the current time zone, thereby improving the intelligent level of traffic management;

[0071] Invention concept B: Based on the intelligent prediction of the number of people flowing in the current time zone, the corresponding number of monitoring resources and management resources are allocated to the current time zone in advance, so that the traffic data of each intersection during the peak time period matches the monitoring and management resources allocated to the intersection, thereby realizing the dynamic allocation of limited monitoring and management resources in the same city.

[0072] The dynamic resource allocation system for traffic sections of the present invention can intelligently predict the number of pedestrians in the current time zone of the traffic section intersection to address the technical problems of difficulty in balancing traffic section congestion and limited resources in the prior art, and then dynamically allocate monitoring resources and management resources for the intersection based on the prediction results, thereby improving the management level of the traffic section.

[0073] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present disclosure, rather than to limit them. Although the present disclosure has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present disclosure, and they should all be included in the scope of the claims and specification of the present disclosure.

Claims

1. A traffic section dynamic resource allocation system, characterized in that: The system comprises: An interval acquisition mechanism is arranged above a certain intersection of a traffic section, and is used to perform continuous image acquisition actions at preset time intervals on the environment of the intersection in a bird's-eye view mode, so as to obtain a plurality of interval acquisition frames in each time partition; A signal analysis mechanism is connected to the interval acquisition mechanism and is used to perform the following signal analysis processing for each time partition: detecting the number of human imaging areas in each interval acquisition frame belonging to the time partition, and performing deduplication processing on each human imaging area in multiple interval acquisition frames within the time partition to obtain the total number of the remaining multiple human imaging areas as the reference number of people corresponding to the time partition; wherein, a plurality of human body imaging regions with matching region edge shapes are treated as the same human body imaging region and deduplication processing is performed to retain only one human body imaging region among the plurality of human body imaging regions for participating in the subsequent accumulation of the total number of the remaining plurality of human body imaging regions; wherein, based on the reference outline of the human body, identification of the human body imaging area in each interval acquisition frame within the time partition is performed; A quantity prediction mechanism is used to, when a time partition belonging to a peak time period of traffic at an intersection is reached on each day, use the arrived time partition as a target time partition, use a plurality of reference numbers of personnel corresponding to a plurality of time partitions of a set number before the target time partition, the coverage area of ​​the intersection, and the total number of traffic sections intersecting at the intersection as a plurality of input data of an artificial intelligence prediction body, so as to run the artificial intelligence prediction body and obtain the number of personnel output by the artificial intelligence prediction body as the reference number of personnel corresponding to the target time partition; A learning execution device, connected to the quantity prediction mechanism, for sending the deep neural network after completing a fixed number of learning operations as an artificial intelligence prediction body to the quantity prediction mechanism for use; A strategy customization device, connected to the quantity prediction mechanism, for dispatching different quantities of monitoring resources and management resources to the intersection based on the reference number of personnel corresponding to the target time partition; Wherein, in the quantity prediction mechanism, the value of the set number is positively correlated with the value of the preset time interval; The method of sending the deep neural network after completing a fixed number of learning operations as an artificial intelligence prediction body to the quantity prediction mechanism for use includes: the smaller the coverage area of ​​the intersection, the smaller the value of the fixed number; The time detection device is connected to the quantity prediction mechanism and is used to determine the arrival of the time partition of the peak time period for traffic at the intersection on a certain day when the current time is equal to the starting time of the time partition of the peak time period for traffic at the intersection on a certain day.

2. The traffic section dynamic resource allocation system according to claim 1, characterized in that: Dispatching different numbers of monitoring resources and management resources to the intersection based on the number of reference personnel corresponding to the target time partition further includes: the more the number of reference personnel corresponding to the target time partition, the more the number of remote management personnel assigned to the intersection; Wherein, dispatching different numbers of monitoring resources and management resources to the intersection based on the number of reference personnel corresponding to the target time partition includes: the more the number of reference personnel corresponding to the target time partition, the more monitoring cameras allocated to the intersection; Among them, the more reference personnel the target time partition corresponds to, the more surveillance cameras are allocated to the intersection, including: one surveillance camera can be shared by multiple intersections, and when the monitoring direction of a surveillance camera is the intersection, it is determined that the surveillance camera is allocated to the intersection.

3. The traffic section dynamic resource allocation system according to claim 2, characterized in that: The system further comprises: A dynamic storage chip is connected to the learning execution device and is used to store various operating parameters of the artificial intelligence prediction body.

4. The traffic section dynamic resource allocation system according to any one of claims 1 to 3, characterized in that: The method of using a bird's-eye view mode to perform continuous image acquisition actions at preset time intervals on the environment where the intersection is located to obtain multiple interval acquisition frames in each time partition includes: adjusting the bird's-eye view angle of the interval acquisition mechanism so that the intersection is included in the bird's-eye view of the interval acquisition mechanism.

5. The traffic section dynamic resource allocation system as claimed in claim 4, characterized in that: Using a bird's-eye view mode to perform continuous image acquisition actions at preset time intervals on the environment where the intersection is located to obtain multiple interval acquisition frames in each time partition also includes: adjusting the bird's-eye view angle of the interval acquisition mechanism so that the bird's-eye view of the interval acquisition mechanism is centered on the intersection.

6. The traffic section dynamic resource allocation system as claimed in claim 5, characterized in that: Adjusting the overhead viewing angle of the interval acquisition mechanism so that the overhead viewing field of the interval acquisition mechanism takes the intersection as the viewing center includes: the centroid position of the imaging area of ​​the intersection in each interval acquisition frame is at the center position of each interval acquisition frame.

7. The traffic section dynamic resource allocation system according to claim 1, characterized in that: Deduplication processing is performed on each human imaging area in the multiple interval acquisition frames within the time partition to obtain the total number of the remaining multiple human imaging areas and use it as the number of reference personnel corresponding to the time partition, which also includes: each human imaging area in the multiple interval acquisition frames within the time partition comes from a certain interval acquisition frame among the multiple interval acquisition frames.

Images