CN-121980411-A - Subway carriage congestion degree monitoring method, system, equipment and medium

CN121980411ACN 121980411 ACN121980411 ACN 121980411ACN-121980411-A

Abstract

The application relates to the technical field of intelligent traffic systems, in particular to a subway carriage congestion degree monitoring method, system, equipment and medium, which comprises the steps of collecting image data, carriage weight data and environmental parameter data in a carriage in real time; the method comprises the steps of carrying out dynamic tracking on passengers based on image data to obtain passenger counts in a carriage, respectively carrying out normalization processing on passenger counts, calculated estimated passenger quantity values based on weight data and environment parameter data to obtain a visual density index, a weight density index and an environment density index, calculating a comprehensive crowding degree index through a weighted fusion formula, and determining the crowding degree grade of the carriage based on the comprehensive crowding degree index. The method overcomes the limitation of a single data source by fusing multi-source sensing data, and realizes real-time, accurate and comprehensive assessment of the congestion degree of the subway carriage.

Inventors

LIU DEKUN
SUN CHAO
JIANG QING
LI XINGCHEN
WANG GUIXIANG
GENG QIWEI

Assignees

浪潮智慧科技有限公司

Dates

Publication Date: 20260505
Application Date: 20251031

Claims (10)

1. The method for monitoring the congestion degree of the subway carriage is characterized by comprising the following steps of: S1, acquiring multi-source sensing data in a subway carriage in real time, wherein the multi-source sensing data comprise image data, carriage weight data and environment parameter data, and the environment parameter data comprise temperature and humidity; s2, carrying out dynamic passenger tracking based on the image data to obtain passenger count in the carriage; S3, carrying out normalization processing on the passenger count in the carriage to obtain a visual density index; calculating a passenger number estimated value based on the carriage weight data, and then carrying out normalization processing to obtain a weight density index; Carrying out normalization processing on the environmental parameter data to obtain an environmental density index; S4, calculating a comprehensive crowdedness index based on the visual density index, the weight density index and the environment density index through weighted fusion, wherein the formula is as follows: Wherein, the Representing an integrated congestion degree index; Representing a visual density index; Represents a weight density index; Representing an environmental density index; representing a preset passenger count weight; Representing a preset passenger weight; representing preset environmental parameter weights; And S5, determining the congestion degree grade of the subway carriage based on the comprehensive congestion degree index.
2. The subway car congestion level monitoring method according to claim 1, wherein in step S1, the image data is specifically video stream data; in step S2, based on the image data, passenger dynamic tracking is performed by adopting YOLOv to 4 target detection algorithm and DeepSORT multi-target tracking algorithm, so as to obtain passenger count in the carriage.
3. The subway car congestion level monitoring method according to claim 1, wherein step S2 specifically includes: S201, extracting image frames according to set frequency from image data input in a video stream mode, carrying out gray correction and noise suppression on each image frame, and cutting out invalid areas through ROI areas to obtain preprocessed image frames in time sequence; S202, inputting each preprocessed image frame into a pre-trained passenger target detection model, identifying each passenger target and generating boundary frame information of each passenger target, wherein the boundary frame information at least comprises coordinates and confidence of the boundary frame in the current image frame, and the passenger target detection model is constructed based on YOLOv target detection algorithm; s203, inputting DeepSORT the preprocessed image frames and the corresponding bounding box information into a multi-target tracking algorithm in sequence according to time sequence for processing, wherein the method comprises the following steps: S2031, according to the coordinates of the boundary frame, intercepting a corresponding image area from the current image frame, and extracting appearance characteristics of passengers in the area; s2032, predicting the position of each track in the current image frame by using Kalman filtering based on the existing passenger tracking track set, and carrying out association matching on the boundary frame of the image frame and the track of the predicted position by using a Hungary algorithm; Wherein the passenger tracking track set is an empty set when processing an initial image frame of the video stream and is dynamically updated as the image frames input in time sequence are processed; S2033. According to the result of the association matching, the following processing is performed: for a successfully matched boundary box and track, updating the corresponding track by using the position and appearance characteristics of the boundary box; for a bounding box that does not match any tracks, creating a new tracking track for it and assigning a new passenger ID; For tracks that do not match any bounding box, marking them as transient and deleting after consecutive frames do not match; s2034, outputting a tracking result list aiming at the current image frame and updating a passenger tracking track set, wherein the tracking result list comprises all tracking tracks in an activated state and corresponding passenger IDs and boundary frame coordinates; S204, counting the number of passengers in the subway carriage based on the tracking result list corresponding to the latest image frame, and obtaining the passenger count in the carriage.
4. The method for monitoring the congestion degree of a subway car according to claim 1, wherein in the step S3, the formula for normalizing the passenger count in the subway car by using the min-max normalization is: Wherein, the Representing a visual density index; representing passenger counts in the vehicle cabin; Representing the number of the fixed personnel of the subway carriage; 。
5. The subway car congestion level monitoring method according to claim 1, wherein in step S3, the formula for calculating the estimated passenger number value based on the car weight data is: Wherein, the A passenger number estimate; Representing car weight data; the weight data of the carriage when no load exists; indicating the average weight of the passengers, ; The formula for normalizing the estimated number of passengers is: Wherein, the Represents a weight density index; Representing the number of the fixed personnel of the subway carriage; 。
6. The method for monitoring the congestion level of a subway car according to claim 1, wherein in step S3, the formula for normalizing the environmental parameter data is as follows Wherein, the Representing an environmental density index; a dimensionless number representing the ambient temperature in the ambient parameter data; a dimensionless number representing the highest allowable ambient celsius temperature in the subway car, ; Representing the relative humidity in the environmental parameter data, ; Indicating the theoretical maximum value of the relative humidity, ; Representing the environmental factor influence factor when meeting Or (b) In the case of any one of the above, Otherwise 。
7. The method for monitoring the congestion degree of the subway carriage according to claim 1, further comprising the step S6 of generating a visual instruction based on the congestion degree level of the subway carriage and sending the visual instruction to a display terminal for visual display; The visual instruction is packaged in a JSON format and comprises the following fields of a subway train ID, a subway carriage number, a current station, a next station, a congestion level, instruction generation time and instruction display time.
8. A subway car congestion degree monitoring system, configured to implement the subway car congestion degree monitoring method according to any one of claims 1 to 7, comprising: The multi-source sensing data acquisition module is used for acquiring multi-source sensing data in a subway carriage in real time, wherein the multi-source sensing data comprise image data, carriage weight data and environment parameter data, and the environment parameter data comprise temperature and humidity; the passenger counting module in the carriage is used for dynamically tracking passengers based on the image data to obtain the passenger count in the carriage; the multi-source data processing module is used for carrying out normalization processing on the passenger count in the carriage to obtain a visual density index; calculating a passenger number estimated value based on the carriage weight data, and then carrying out normalization processing to obtain a weight density index; Carrying out normalization processing on the environmental parameter data to obtain an environmental density index; the comprehensive crowding degree index calculation module is used for calculating the comprehensive crowding degree index through weighted fusion based on the visual density index, the weight density index and the environment density index; and the congestion degree grade determining module is used for determining the congestion degree grade of the subway carriage based on the comprehensive congestion degree index.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor is configured to implement the steps of the subway car congestion level monitoring method of any one of claims 1-7 when the computer program is executed.
10. A storage medium having a computer program stored thereon, which when executed by a processor performs the steps of the subway car congestion degree monitoring method according to any one of claims 1 to 7.

Description

Subway carriage congestion degree monitoring method, system, equipment and medium Technical Field The application relates to the technical field of intelligent traffic systems, in particular to a subway carriage congestion degree monitoring method, system, equipment and medium. Background With the rapid development of urban rail transit, subways are taken as main public transportation modes, and the operation efficiency and the service quality of the subways are increasingly focused. The degree of congestion of the subway carriage directly influences the riding experience and operation safety of passengers, so that the real-time monitoring of the degree of congestion of the carriage and guiding of the passengers to balance riding become an important research direction in the field of intelligent transportation. In the prior art, a method based on sensor data is generally adopted to monitor the congestion degree of a subway carriage. For example, the number of passengers in a carriage is estimated by counting the number of passengers on and off a carriage door by installing an infrared sensor or a laser counter, the number of passengers is counted by using a monitoring system to collect images in the carriage through a camera and using a target detection algorithm, and the number of passengers is estimated by measuring weight change by using a pressure sensor at the bottom of the carriage. The methods can reflect the congestion degree of the subway carriage to a certain extent. However, the prior art has the defects that the influence of environmental interference or measurement errors is easily caused by relying on a single data source, such as the influence of complicated light in a carriage, shielding among passengers and the like is easily caused by relying on image recognition, and the weight fluctuation and the like caused by the movement of passengers, the placement of goods and the like are difficult to distinguish by relying on weight data. The dependence on a single data source causes the accuracy and the anti-interference capability of the congestion degree estimation result to be insufficient, and the deviation is easy to occur in a dynamic and changing actual operation environment. Disclosure of Invention Aiming at the technical problems that the existing carriage congestion degree monitoring method mainly relies on a single type of data source to judge the carriage congestion degree, so that an estimation result is easy to be interfered by a specific mode and the accuracy and the robustness are insufficient, the application provides the subway carriage congestion degree monitoring method, system, equipment and medium, which are used for comprehensively calculating by fusing multi-source data such as images, weight, environmental parameters and the like, and effectively improving the accuracy and the reliability of carriage congestion degree estimation. In a first aspect, the present application provides a method for monitoring the congestion degree of a subway carriage, comprising the following steps: S1, acquiring multi-source sensing data in a subway carriage in real time, wherein the multi-source sensing data comprise image data, carriage weight data and environment parameter data, and the environment parameter data comprise temperature and humidity; s2, carrying out dynamic passenger tracking based on the image data to obtain passenger count in the carriage; S3, carrying out normalization processing on the passenger count in the carriage to obtain a visual density index; calculating a passenger number estimated value based on the carriage weight data, and then carrying out normalization processing to obtain a weight density index; Carrying out normalization processing on the environmental parameter data to obtain an environmental density index; S4, calculating a comprehensive crowdedness index based on the visual density index, the weight density index and the environment density index through weighted fusion, wherein the formula is as follows: Wherein, the Representing an integrated congestion degree index; Representing a visual density index; Represents a weight density index; Representing an environmental density index; representing a preset passenger count weight; Representing a preset passenger weight; representing preset environmental parameter weights; And S5, determining the congestion degree grade of the subway carriage based on the comprehensive congestion degree index. In step S1, image data is collected by a wide-angle camera disposed at the top of the subway carriage; collecting carriage weight data through a pressure sensor arranged at the bottom of a subway carriage; Environmental parameter data are collected through a temperature and humidity sensor arranged in a subway carriage. It should be further noted that, in step S1, the image data is specifically video stream data; in step S2, based on the image data, passenger dynamic tracking is performed by adopting YOLOv to 4 ta