CN112232333A - Real-time passenger flow thermodynamic diagram generation method in subway station - Google Patents
Real-time passenger flow thermodynamic diagram generation method in subway station Download PDFInfo
- Publication number
- CN112232333A CN112232333A CN202011499484.0A CN202011499484A CN112232333A CN 112232333 A CN112232333 A CN 112232333A CN 202011499484 A CN202011499484 A CN 202011499484A CN 112232333 A CN112232333 A CN 112232333A
- Authority
- CN
- China
- Prior art keywords
- real
- passenger flow
- density
- subway station
- thermodynamic diagram
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/41—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformation in the plane of the image
- G06T3/40—Scaling the whole image or part thereof
- G06T3/4038—Scaling the whole image or part thereof for image mosaicing, i.e. plane images composed of plane sub-images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
- G06V20/53—Recognition of crowd images, e.g. recognition of crowd congestion
-
- G—PHYSICS
- G09—EDUCATION; CRYPTOGRAPHY; DISPLAY; ADVERTISING; SEALS
- G09B—EDUCATIONAL OR DEMONSTRATION APPLIANCES; APPLIANCES FOR TEACHING, OR COMMUNICATING WITH, THE BLIND, DEAF OR MUTE; MODELS; PLANETARIA; GLOBES; MAPS; DIAGRAMS
- G09B29/00—Maps; Plans; Charts; Diagrams, e.g. route diagram
Abstract
The invention discloses a method for generating a real-time passenger flow thermodynamic diagram in a subway station, which belongs to the technical field of computer vision and intelligent monitoring. The invention utilizes the monitoring camera in the subway station to generate the real-time passenger flow thermodynamic diagram, is convenient for management personnel in the subway station to observe the real-time distribution and change conditions of passenger flow, and is beneficial to safe operation and refined passenger flow management of the subway.
Description
Technical Field
The invention belongs to the technical field of computer vision and intelligent monitoring, and particularly relates to a real-time passenger flow thermodynamic diagram generation method in a subway station.
Background
With the development of urban rail transit, especially the rapid development of subways in various cities, the traveling modes of people are also changing. Because urban roads are very congested in rush hours, more and more urban crowds choose to take subways to go out, and higher requirements are provided for passenger flow operation and management in subway stations.
At present, an effective passenger flow monitoring mode is lacked in a subway station, the current passenger flow condition can be judged only through limited camera monitoring pictures and rich experience of subway transportation and management personnel, the passenger flow distribution condition of the whole station in the subway cannot be accurately obtained, and more refined passenger flow management and evacuation cannot be carried out.
With the development of artificial intelligence and machine vision technology, effective passenger flow monitoring and auxiliary passenger flow management can be carried out by utilizing a monitoring camera in the subway station. The monitoring cameras at the positions of an inlet, an outlet, a passage, a station hall, a platform, an escalator and the like in the subway station are utilized to monitor the distribution and the change condition of the passenger flow density of the whole station in real time, and the passenger flow density of each area is estimated by utilizing artificial intelligence and machine vision technology, so that the method is helpful for assisting managers in the subway station to finely manage the passenger flow.
Disclosure of Invention
The purpose of the invention is as follows: the invention aims to provide a method for generating a real-time passenger flow thermodynamic diagram in a subway station, aiming at the problem that the real-time passenger flow density distribution and the change condition are not effectively monitored in the subway station.
The technical scheme is as follows: in order to achieve the purpose, the invention adopts the following technical scheme:
a real-time passenger flow thermodynamic diagram generation method in a subway station comprises the following steps:
step 4, reading a real-time RTSP video stream of a monitoring camera in the station, and decoding the real-time RTSP video stream into a picture frame;
step 5, aiming at each frame of picture, carrying out density estimation by using a trained convolutional neural network model, and generating a thermodynamic diagram;
and 6, splicing the thermodynamic diagrams of each area into a whole station passenger flow density diagram, setting an alarm threshold value for each area, and generating an alarm event to remind attention when the passenger flow density of the area exceeds the threshold value.
Further, in step 1, when the camera is selected, the selection condition includes an installation angle and a monitoring range of the camera.
3. The method for generating the real-time passenger flow thermodynamic diagram in the subway station according to claim 1, wherein: in the step 2, the data labeling method is as follows:
a）p i is the true coordinates of the object in the image,prepresenting the coordinates of the object detected by the object detection model,Hrepresents the sum of error values of coordinate values obtained by all target detections and the true value,Mwhich indicates the number of the detection targets,Hcalculated by the following way:
b) by using a Gaussian kernelG σ Convolving the above equation, i.e. converting it into a continuous density function, the density map F is represented as follows:
further, the convolutional neural network model in step 3 takes VGG16 as a front-end model, and adopts ordinary convolution and hole convolution to design a back-end model, and finally trains the model by using the following loss function:
wherein, theta represents a network parameter,Nrepresenting the number of the training pictures,X i is the input of a picture or a picture,F M is a density map of the input image estimated by the network model,M ithe table enters a true density map of the image,F C is the number of people estimated by the network model of the input image,C i representing the actual number of people who input the image,L M is the deviation value between the density map estimated by the network model of the input image and the real density map,L C is the deviation value between the estimated number of people and the actual number of people of the input image through the network model.
Further, in the step 4, the decoding mode adopts an NVCODEC hard solution mode, which includes the following steps:
firstly, initializing a CUVID hard solution module;
and then acquiring video stream data by calling an interface in an FFMPEG library, sending the data to a hard solution module, and converting the YUV image obtained by hard solution into an RGB image by using a CUDA if the hard solution is successful. And if the hard decoding is unsuccessful, the CUVID hard decoding module is reinitialized, and the step 4 is repeated.
Further, in the step 5, a Caffe deep learning framework and a GPU acceleration are adopted for a model of density estimation, the model of density estimation is obtained by training through the Caffe framework, and the framework supports NVIDIA GPU-based calculation acceleration.
Has the advantages that: compared with the prior art, the method for generating the real-time passenger flow thermodynamic diagram in the subway station comprises the steps of selecting a suitable real-time monitoring video of a camera through a plane diagram of the subway station and the distribution of the camera in the station, collecting the monitoring pictures in each scene and generating a marking sample, training a convolutional neural network model, reading a real-time RTSP video stream and decoding the RTSP video stream into a picture, inputting a passenger flow density analysis model to obtain the real-time passenger flow thermodynamic diagram of each area, splicing the passenger flow thermodynamic diagrams of the areas such as an inlet, an outlet, a channel, a station hall, a station platform, an escalator and the like in the station to generate the whole station real-time passenger flow thermodynamic diagram, and generating alarm information when the passenger flow density of a certain area exceeds a set threshold value, sending a corresponding alarm event to a station service system to remind a station service staff of paying. The invention utilizes the monitoring camera in the subway station to generate the real-time passenger flow thermodynamic diagram, is convenient for management personnel in the subway station to observe the real-time distribution and change conditions of passenger flow, and is beneficial to safe operation and refined passenger flow management of the subway.
Drawings
FIG. 1 is a system architecture diagram of a method for generating a real-time passenger flow thermodynamic diagram in a subway station;
FIG. 2 is a labeling diagram;
FIG. 3 is a real scene diagram of a subway
FIG. 4 is a schematic diagram of thermodynamic diagram generation corresponding to a subway live-action map;
FIG. 5 is a diagram of a crowd density estimation model architecture;
fig. 6 is a hard solution flow chart.
Detailed Description
The present invention will be further described with reference to the following embodiments.
A real-time passenger flow thermodynamic diagram generation method in a subway station comprises the following steps:
step 1: selecting a proper monitoring video according to a plane map in the subway station and the installation position of a monitoring camera, wherein the selected position comprises key passenger flow areas such as station entrances and exits, channels, escalators, station halls, platforms and the like;
step 2: selecting a monitoring camera with a proper scene and collecting pictures, and selecting a proper amount of pictures for data annotation in each scene;
and step 3: training a convolutional neural network model by using the labeled pictures;
and 4, step 4: reading a real-time RTSP video stream of a monitoring camera in a station, and decoding the real-time RTSP video stream into a picture frame;
and 5: for each frame of picture, performing density estimation by using a trained convolutional neural network model, and generating a thermodynamic diagram;
step 6: and splicing the thermodynamic diagrams of each area into a whole station passenger flow density diagram, setting an alarm threshold value for each area, and generating an alarm event when the passenger flow density of the area exceeds the threshold value to remind the station staff of paying attention.
In the step 1, when the camera is selected, the installation angle and the monitoring range of the camera need to be considered, so that people in the monitoring view screen are shielded less, and all people in the view range can be seen as far as possible.
In step 2, the image annotation method generates a thermodynamic diagram by using the following two formulas:
a）p i is the true coordinates of the object in the image,prepresenting the coordinates of the object detected by the object detection model,Hrepresents the sum of error values of coordinate values obtained by all target detections and the true value,Mwhich indicates the number of the detection targets,Hcalculated by the following way:
b) by using a Gaussian kernelG σ Convolving the above equation, i.e. converting it into a continuous density function, the density map F is represented as follows:
the convolutional neural network model in the step 3 takes VGG16 as a front-end model, adopts common convolution and hole convolution to design a rear-end model, and finally trains the model by using the following loss function:
wherein, theta represents a network parameter,Nrepresenting the number of the training pictures,X i is the input of a picture or a picture,F M is inputting images via networkThe density map estimated by the model is,M ithe table enters a true density map of the image,F C is the number of people estimated by the network model of the input image,C i representing the actual number of people who input the image,L M is the deviation value between the density map estimated by the network model of the input image and the real density map,L C is the deviation value between the estimated number of people and the actual number of people of the input image through the network model.
The decoding mode in step 4 adopts an NVCODEC (NvidiaCodec) hard solution mode, firstly, a CUVID hard solution module is initialized, then, the interface in the FFMPEG library is called to obtain video stream data, the data is sent to the hard solution module, and if the hard solution is successful, a CUDA (computer Unified Device architecture) is used to convert the YUV image obtained by the hard solution into an RGB image. And if the hard decoding is unsuccessful, the CUVID hard decoding module is reinitialized, and the step 4 is repeated.
The density analysis model in the step 5 is accelerated by adopting a Caffe deep learning framework and a GPU, the density analysis model is obtained by training through the Caffe framework, and the framework can support calculation acceleration based on the NVIDIA GPU.
Example 1:
the method for generating the real-time passenger flow thermodynamic diagram in the subway station comprises the steps of selecting a real-time monitoring video of a proper camera through a plane diagram of the subway station and distribution of the cameras in the station, collecting monitoring pictures in various scenes and generating a labeling sample, training a convolutional neural network model, reading a real-time RTSP video stream and decoding the RTSP video stream into pictures, inputting a passenger flow density analysis model to obtain the real-time passenger flow thermodynamic diagram of each area, splicing the passenger flow thermodynamic diagrams of the areas such as an inlet, an outlet, a channel, a station hall, a station platform, a staircase and the like in the station to generate the whole-station real-time passenger flow thermodynamic diagram, and generating alarm information when the passenger flow density of a certain area exceeds a set threshold value, sending a corresponding alarm event to a station service system and reminding station service staff of paying attention. The specific process is as follows:
step 1: and selecting a proper monitoring video according to the plane map in the subway station and the installation position of the monitoring camera, wherein the selected position comprises key passenger flow areas such as station entrances and exits, channels, escalators, station halls, platforms and the like.
The system architecture is shown in fig. 1.
When the camera is selected, the installation angle and the monitoring range of the camera need to be considered, so that people in the monitoring screen are shielded less, all people in the visual field range can be seen as far as possible, the selected installation angle of the camera is a horizontal included angle of 30-75 degrees, and the effective monitoring distance is 0-20 meters.
Step 2: and selecting a monitoring camera with a proper scene and acquiring pictures, and selecting a proper amount of pictures for data annotation in each scene.
The labeling tool is written by matlab, and the data labeling example is shown in FIG. 2.
The image annotation method adopts the following two formulas to generate a thermodynamic diagram:
a) assume the location of the annotation point isp i Then a label with N headers can be represented as follows:
b) by using a Gaussian kernelG σ Convolving the above equation, i.e. converting it into a continuous density function, the density map can be represented as follows:
fig. 3 shows a subway live-action map, and a corresponding thermodynamic diagram generation schematic diagram is shown in fig. 4.
And step 3: the labeled pictures are used to train a convolutional neural network model.
The convolutional neural network is VGG16, the model structure is shown in FIG. 5, and the loss function adopted by training is:
and 4, step 4: and reading the real-time RTSP video stream of the monitoring camera in the station, and decoding the real-time RTSP video stream into picture frames.
The hard solution flow is shown in fig. 6.
And 5: and performing density estimation by using a trained convolutional neural network model for each frame of picture, and generating a thermodynamic diagram.
The model adopts Caffe deep learning framework and uses NVIDIA RTX2080Ti and CUDA to accelerate the depth model reasoning speed.
Step 6: and splicing the thermodynamic diagrams of each area into a whole station passenger flow density diagram, setting an alarm threshold value for each area, and generating an alarm event when the passenger flow density of the area exceeds the threshold value to remind the station staff of paying attention.
The above description is only a preferred embodiment of the present invention, and it will be apparent to those skilled in the art that various modifications and variations can be made without departing from the technical principles of the present invention, and these modifications and variations should also be construed as the scope of the present invention.
Claims (5)
1. A real-time passenger flow thermodynamic diagram generation method in a subway station is characterized by comprising the following steps: the method comprises the following steps:
step 1, selecting a proper monitoring video according to a plane map in a subway station and the installation position of a monitoring camera;
step 2, selecting a monitoring camera of a proper scene, collecting pictures, and performing data annotation on the pictures selected in each scene;
step 3, training a convolutional neural network model by using the labeled pictures;
step 4, reading a real-time RTSP video stream of a monitoring camera in the station, and decoding the real-time RTSP video stream into a picture frame;
step 5, aiming at each frame of picture, carrying out density estimation by using a trained convolutional neural network model, and generating a thermodynamic diagram;
step 6, splicing the thermodynamic diagrams of each area into a whole station passenger flow density diagram, setting an alarm threshold value for each area, and generating an alarm event when the passenger flow density of the area exceeds the threshold value;
in the step 2, the data labeling method is as follows:
a）p i is the true coordinates of the object in the image,prepresenting the coordinates of the object detected by the object detection model,Hrepresents the sum of error values of coordinate values obtained by all target detections and the true value,Mwhich indicates the number of the detection targets,Hcalculated by the following way:
b) by using a Gaussian kernelG σ Convolving the above equation, i.e. converting it into a continuous density function, the density map F is represented as follows:
2. the method for generating the real-time passenger flow thermodynamic diagram in the subway station according to claim 1, wherein: in the step 1, when the camera is selected, the selection condition includes an installation angle and a monitoring range of the camera.
3. The method for generating the real-time passenger flow thermodynamic diagram in the subway station according to claim 2, wherein: the convolutional neural network model in the step 3 takes VGG16 as a front-end model, adopts common convolution and hole convolution to design a rear-end model, and finally trains the model by using the following loss function:
wherein, theta represents a network parameter,Nrepresenting the number of the training pictures,X i is the input of a picture or a picture,F M is a density map of the input image estimated by the network model,M ithe table enters a true density map of the image,F C is an inputThe number of people the image is estimated to have through the network model,C i representing the actual number of people who input the image,L M is the deviation value between the density map estimated by the network model of the input image and the real density map,L C is the deviation value between the estimated number of people and the actual number of people of the input image through the network model.
4. The method for generating the real-time passenger flow thermodynamic diagram in the subway station according to claim 1, wherein: in the step 4, the decoding mode adopts an NVCODEC hard solution mode, and the method comprises the following steps:
firstly, initializing a CUVID hard solution module;
and then acquiring video stream data by calling an interface in an FFMPEG library, sending the data to a hard solution module, and converting the YUV image obtained by hard solution into an RGB image by using a CUDA if the hard solution is successful.
5. The method for generating the real-time passenger flow thermodynamic diagram in the subway station according to claim 1, wherein: in the step 5, the density estimation model is accelerated by adopting a Caffe deep learning framework and a GPU, the density estimation model is obtained by using the Caffe framework for training, and the framework supports NVIDIA GPU-based calculation acceleration.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011499484.0A CN112232333A (en) | 2020-12-18 | 2020-12-18 | Real-time passenger flow thermodynamic diagram generation method in subway station |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011499484.0A CN112232333A (en) | 2020-12-18 | 2020-12-18 | Real-time passenger flow thermodynamic diagram generation method in subway station |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112232333A true CN112232333A (en) | 2021-01-15 |
Family
ID=74124907
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011499484.0A Pending CN112232333A (en) | 2020-12-18 | 2020-12-18 | Real-time passenger flow thermodynamic diagram generation method in subway station |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112232333A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113343960A (en) * | 2021-08-06 | 2021-09-03 | 南京信息工程大学 | Method for estimating and early warning passenger flow retained in subway station in real time |
| CN113793400A (en) * | 2021-09-14 | 2021-12-14 | 南京信息工程大学 | Construction method of gas concentration thermodynamic diagram |
| CN113807026A (en) * | 2021-10-08 | 2021-12-17 | 青岛理工大学 | Passenger flow line optimization and dynamic guide signboard system in subway station and design method |
| CN114170568A (en) * | 2021-12-03 | 2022-03-11 | 成都鼎安华智慧物联网股份有限公司 | Personnel density detection method and system based on deep learning |
| CN114332778A (en) * | 2022-03-08 | 2022-04-12 | 深圳市万物云科技有限公司 | Intelligent alarm work order generation method and device based on people stream density and related medium |
| CN114581846A (en) * | 2022-03-03 | 2022-06-03 | 北京城建设计发展集团股份有限公司 | Method and device for monitoring holographic passenger flow of subway station in real time and computer equipment |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105549727A (en) * | 2015-08-29 | 2016-05-04 | 宇龙计算机通信科技(深圳)有限公司 | Visitor flow rate reminding method and device |
| CN109815867A (en) * | 2019-01-14 | 2019-05-28 | 东华大学 | A kind of crowd density estimation and people flow rate statistical method |
| CN110245579A (en) * | 2019-05-24 | 2019-09-17 | 北京百度网讯科技有限公司 | Density of stream of people prediction technique and device, computer equipment and readable medium |
-
2020
- 2020-12-18 CN CN202011499484.0A patent/CN112232333A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105549727A (en) * | 2015-08-29 | 2016-05-04 | 宇龙计算机通信科技(深圳)有限公司 | Visitor flow rate reminding method and device |
| CN109815867A (en) * | 2019-01-14 | 2019-05-28 | 东华大学 | A kind of crowd density estimation and people flow rate statistical method |
| CN110245579A (en) * | 2019-05-24 | 2019-09-17 | 北京百度网讯科技有限公司 | Density of stream of people prediction technique and device, computer equipment and readable medium |
Non-Patent Citations (1)
| Title |
|---|
| YANJIE WANG等: "Multi-scale dilated convolution of convolutional neural network for crowd counting", 《HTTPS://LINK.SPRINGER.COM/ARTICLE/10.1007/S11042-019-08208-6》 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113343960A (en) * | 2021-08-06 | 2021-09-03 | 南京信息工程大学 | Method for estimating and early warning passenger flow retained in subway station in real time |
| CN113793400A (en) * | 2021-09-14 | 2021-12-14 | 南京信息工程大学 | Construction method of gas concentration thermodynamic diagram |
| CN113807026A (en) * | 2021-10-08 | 2021-12-17 | 青岛理工大学 | Passenger flow line optimization and dynamic guide signboard system in subway station and design method |
| CN114170568A (en) * | 2021-12-03 | 2022-03-11 | 成都鼎安华智慧物联网股份有限公司 | Personnel density detection method and system based on deep learning |
| CN114581846A (en) * | 2022-03-03 | 2022-06-03 | 北京城建设计发展集团股份有限公司 | Method and device for monitoring holographic passenger flow of subway station in real time and computer equipment |
| CN114581846B (en) * | 2022-03-03 | 2024-02-20 | 北京城建设计发展集团股份有限公司 | Real-time monitoring method and device for holographic passenger flow of subway station and computer equipment |
| CN114332778A (en) * | 2022-03-08 | 2022-04-12 | 深圳市万物云科技有限公司 | Intelligent alarm work order generation method and device based on people stream density and related medium |
| CN114332778B (en) * | 2022-03-08 | 2022-06-21 | 深圳市万物云科技有限公司 | Intelligent alarm work order generation method and device based on people stream density and related medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112232333A (en) | Real-time passenger flow thermodynamic diagram generation method in subway station | |
| CN107004271B (en) | Display method, display apparatus, electronic device, computer program product, and storage medium | |
| CN103795976B (en) | A kind of full-time empty 3 d visualization method | |
| CN106203513A (en) | A kind of based on pedestrian's head and shoulder multi-target detection and the statistical method of tracking | |
| CN104092988A (en) | Method, device and system for managing passenger flow in public place | |
| KR20080085837A (en) | Object density estimation in vedio | |
| CN106373430A (en) | Intersection pass early warning method based on computer vision | |
| Jin et al. | Adapt: Action-aware driving caption transformer | |
| CN102759347A (en) | Online in-process quality control device and method for high-speed rail contact networks and composed high-speed rail contact network detection system thereof | |
| CN116824859B (en) | Intelligent traffic big data analysis system based on Internet of things | |
| CN107480653A (en) | passenger flow volume detection method based on computer vision | |
| CN113159004B (en) | Passenger flow estimation method for rail transit carriage | |
| CN111767798A (en) | Intelligent broadcasting guide method and system for indoor networking video monitoring | |
| CN113807026A (en) | Passenger flow line optimization and dynamic guide signboard system in subway station and design method | |
| CN112381043A (en) | Flag detection method | |
| CN108471497A (en) | A kind of ship target real-time detection method based on monopod video camera | |
| CN104616277B (en) | Pedestrian's localization method and its device in video structural description | |
| CN110176022A (en) | A kind of tunnel overall view monitoring system and method based on video detection | |
| CN113343960A (en) | Method for estimating and early warning passenger flow retained in subway station in real time | |
| CN114640807B (en) | Video-based object statistics method, device, electronic equipment and storage medium | |
| WO2023070955A1 (en) | Method and apparatus for detecting tiny target in port operation area on basis of computer vision | |
| CN112509190B (en) | Subway vehicle section passenger flow statistical method based on shielded gate passenger flow counting | |
| CN115512263A (en) | Dynamic visual monitoring method and device for falling object | |
| Liu et al. | Metro passenger flow statistics based on yolov3 | |
| CN110287897B (en) | Rail train visual positioning system based on deep learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210115 |