CN110855937A - Error control method of subway station hall people counting system - Google Patents
Error control method of subway station hall people counting system Download PDFInfo
- Publication number
- CN110855937A CN110855937A CN201910908992.0A CN201910908992A CN110855937A CN 110855937 A CN110855937 A CN 110855937A CN 201910908992 A CN201910908992 A CN 201910908992A CN 110855937 A CN110855937 A CN 110855937A
- Authority
- CN
- China
- Prior art keywords
- camera
- station hall
- people
- compensation
- shot
- 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
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/18—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
- H04N7/181—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast for receiving images from a plurality of remote sources
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Indicating And Signalling Devices For Elevators (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses an error control method of a subway station hall people counting system, wherein a first camera is installed at a gate entrance of a subway station hall, a second camera is installed at a gate exit, a plurality of compensation cameras are installed in the station hall, and the shooting range of the compensation cameras can cover the whole station hall; the number of the people who get into the station shot by the first camera in real time is recorded as A1tThe number of the outbound persons shot by the second camera in real time is recorded as A2tThe total number of people in the station hall shot by the plurality of compensation cameras in real time is recorded A3t(ii) a Error delta A of real-time people counting in station halltComprises the following steps: 1) when A3t>At A0,. DELTA.At＝(|A1t‑A2tI) XK; 2) when A3tWhen the ratio is less than or equal to A0, delta At0; the parameters of the first camera and the second camera are the same, and K is the identification precision of the first camera; a0 is the set value. Through setting up the compensation camera, solved first camera and second camera because the error accumulation problem that the discernment precision produced, improved whole station room real-time people number statistics systemAnd (4) the accuracy of the system.
Description
Technical Field
The invention belongs to the technical field of subway security systems, and particularly relates to an error control method of a subway station hall people counting system.
Background
The subway is widely popular as a quick travel vehicle for people. But the subway can also fail for a variety of reasons, resulting in a short outage. Once the subway is stopped, the number of passengers in the platform and the exhibition hall is greatly increased in a short time, and particularly, during the early peak and the late peak, the accidents such as extrusion, even treading and the like are easily caused by congestion. Aiming at the situation, the subway platform is controlled by adopting a manual intervention mode of workers, and passengers outside the station are dissuaded from entering the station when the number of the passengers in the platform and the exhibition hall is more visually observed. However, the accuracy is low, the efficiency is low and the early warning effect is poor by adopting a visual inspection mode.
Therefore, at present, the number of passengers in a station hall is counted by respectively arranging cameras at a gate entrance and a gate exit of the subway station hall, and when the number of passengers in a platform exceeds a threshold value, the gate entrance is closed in time to prevent congestion. In order to accurately count the number of people, the lens of the camera is preferably vertically arranged. However, because the height of the station hall is low, if the cameras are vertically installed, the shooting range of each camera is greatly reduced, and in order to improve the shooting range, the number of the cameras can be greatly increased, so that the installation is unchanged, the workload is increased, the cost is high, and the attractiveness is influenced. Therefore, cameras at the entrance and exit of the station hall can only be obliquely arranged, but the oblique arrangement causes large statistical errors, for example, passengers wearing special clothing and hats and passengers with short height and blocked by others are not easy to capture and identify. In particular, in the peak period with large pedestrian flow, the error is larger, and the error is always accumulated and cannot be automatically cleared, so that the larger the statistical error is.
Disclosure of Invention
The invention aims to provide an error control method of a subway station hall people counting system, which solves the technical problems that in the prior art, the counting error is large, the error is always accumulated and cannot be automatically cleared because cameras at a gate inlet and a gate outlet of a subway station hall are obliquely arranged, and the larger the counting error is, the more the error is.
In order to solve the technical problems, the invention is realized by the following technical scheme:
a first camera is installed at a gate entrance of a subway station hall, a second camera is installed at a gate exit, a plurality of compensation cameras are installed in the station hall, the first camera, the second camera and the compensation cameras are all obliquely arranged, the shooting range of the first camera can cover the gate entrance, the shooting range of the second camera can cover the gate exit, and the shooting ranges of the compensation cameras can cover the whole station hall;
the number of the people who get into the station shot by the first camera in real time is recorded as A1tThe number of the outbound persons shot by the second camera in real time is recorded as A2tThe total number of people in the station hall shot by the plurality of compensation cameras in real time is recorded A3t；
Error delta A of the station hall real-time people counting systemtComprises the following steps:
1) when A3t>At A0,. DELTA.At＝(|A1t-A2t|)×K；
2) When A3tWhen the ratio is less than or equal to A0, delta At＝0；
The models of the first camera and the second camera are the same, K is the recognition accuracy of pictures or videos shot by the first camera and the second camera, and K is the inherent recognition accuracy obtained through statistics; a0 is the set value.
Preferably, the a0 is any one of 10-30, and the specific value is set and adjusted according to the actual station hall condition.
The recognition accuracy of people in the picture or video shot by the camera cannot be 100%, for example, passengers wearing odd clothes and hats and passengers with short height and being shielded by others are not easy to capture and recognize, so that people counting of the picture or video shot by the camera has an error, the error is equal to the product of the number of people shot and the recognition accuracy, and the more the number of people shot, the larger the error is. Meanwhile, with the increase of time, the statistical error of the number of people in the station hall is larger (the error is equal to the superposition of the error of each previous moment), and particularly, the statistical error of the number of people is larger and larger after people stream peak periods such as on duty and off duty in each day, so that the statistical accuracy of the number of people in the station hall is greatly reduced.
The first camera, the second camera and the compensation camera transmit the shot pictures or videos to the controller in real time, and the controller counts the number of people who get in the station A1 in real time through calculationtThe number of people leaving the station A2tAnd the total number of people A3 in the station hall shot by the compensation camera at the momenttAnd according to A3t、A1tAnd A2tCalculating the people counting error delta A of the people counting system of the station hall at the momentt。
By arranging the compensation camera, when the number of people in the station hall is less than or equal to a set value A0, the number of people in the station hall is very small at the moment, the recognition accuracy of the compensation camera is high, and at the moment, the compensation camera shoots the total number of people in the station hall A3tNamely the total number of people in the station hall, and the statistical error of the number of people in the station hall is determined to be about 0.
When the number of people in the station hall is larger than the set value A0, the number of people in the station hall is large, the installation position of the compensation camera is positioned at the corner, the installation inclination angle is large, and the algorithm is different, so that the recognition accuracy of the pictures or videos shot by the compensation camera is low when the flow of people is large, the error of the number of people shot is large, and the number of people shot by the compensation camera is delta At＝(|A1t-A2tError delta A of real-time people counting system of station hall by using |) xKtThe accuracy is relatively higher.
Through setting up the compensation camera, solved first camera and second camera because the error accumulation problem that the discernment precision produced, improved the precision of whole station room real-time people statistical system.
Preferably, the first camera, the second camera and the compensation camera are the same in model, so that the installation and the subsequent image processing are facilitated.
Preferably, if the station hall is provided with a plurality of entrance gates and exit gates, a first camera is installed at each entrance gate, and a second camera is installed at each exit gate. Some station halls are large and are provided with a plurality of entrance gates and exit gates, A1tThe total number of people entering the station hall at the moment is calculated for all the pictures or videos shot by the first cameras of the plurality of entrance gates after being processed by the controller; a2tAnd the total number of people who get out of the hall at the moment is calculated after the pictures or videos shot by all the second cameras of the plurality of entrance gates are processed by the controller.
Preferably, the first camera, the second camera and the compensation camera are installed at an inclination angle of 30-60 degrees, which is determined by the height of the station hall.
Preferably, the compensation cameras are installed at corners of the station hall, shooting ranges of the compensation cameras can cover the whole station hall, passengers standing at the corners are prevented from being missed to shoot, and statistical accuracy is improved.
Compared with the prior art, the invention has the beneficial effects that:
through setting up the compensation camera, solved first camera and second camera because the error accumulation problem that the discernment precision produced, improved the precision of whole station room real-time people statistical system.
Detailed Description
The present invention will be described in further detail with reference to specific embodiments below:
a first camera is installed at a gate entrance of a subway station hall, a second camera is installed at a gate exit, a plurality of compensation cameras are installed in the station hall, the first camera, the second camera and the compensation cameras are all obliquely arranged, the shooting range of the first camera can cover the gate entrance, the shooting range of the second camera can cover the gate exit, and the shooting ranges of the compensation cameras can cover the whole station hall;
the number of the people who get into the station shot by the first camera in real time is recorded as A1tThe number of the outbound persons shot by the second camera in real time is recorded as A2tThe total number of people in the station hall shot by the plurality of compensation cameras in real time is recorded A3t；
Error delta A of the station hall real-time people counting systemtComprises the following steps:
1) when A3t>At A0,. DELTA.At＝(|A1t-A2t|)×K；
2) When A3tWhen the ratio is less than or equal to A0, delta At＝0；
The models of the first camera and the second camera are the same, and K is the recognition accuracy of pictures or videos shot by the first camera and the second camera; a0 is the set value.
In this embodiment, a0 is 20. In other embodiments, a0 may be any one of 10-30, and the specific value is set and adjusted according to the actual station hall situation.
The recognition accuracy of people in the picture or video shot by the camera cannot be 100%, for example, passengers wearing odd clothes and hats and passengers with short height and being shielded by others are not easy to capture and recognize, so that people counting of the picture or video shot by the camera has an error, the error is equal to the product of the number of people shot and the recognition accuracy, and the more the number of people shot, the larger the error is. Meanwhile, with the increase of time, the statistical error of the number of people in the station hall is larger (the error is equal to the superposition of the error of each previous moment), and particularly, the statistical error of the number of people is larger and larger after people stream peak periods such as on duty and off duty in each day, so that the statistical accuracy of the number of people in the station hall is greatly reduced.
The first camera, the second camera and the compensation camera transmit the shot pictures or videos to the controller in real time, and the controller counts the number of people who get in the station A1 in real time through calculationtThe number of people leaving the station A2tAnd the total number of people A3 in the station hall shot by the compensation camera at the momenttAnd according to A3t、A1tAnd A2tCalculating the people counting error delta A of the people counting system of the station hall at the momentt。
By arranging the compensation camera, when the number of people in the station hall is less than or equal to the set value 20, the number of people in the station hall is very small at the moment, the recognition accuracy of the compensation camera is high, and at the moment, the compensation camera shoots the total number of people A3 in the station halltNamely the total number of people in the station hall, and the statistical error of the number of people in the station hall is determined to be about 0.
When the number of people in the station hall is more than the set value 20 (assuming that the current calibration value is 20), the number of people in the station hall is more, because the installation position of the compensation camera is positioned at the corner, the installation inclination angle is larger, and the algorithms are different, the recognition accuracy of the pictures or videos shot by the compensation camera is lower when the flow of people is larger, the error of the number of people shot is larger, and therefore the delta A is used at the momentt＝(|A1t-A2tError delta A of real-time people counting system of station hall by using |) xKtThe accuracy is relatively higher.
Through setting up the compensation camera, solved first camera and second camera because the error accumulation problem that the discernment precision produced, improved the precision of whole station room real-time people statistical system.
In this embodiment, the first camera, the second camera and the compensation camera are of the same model, and the specific model is DS-2CD 2626F.
In other embodiments, the compensation camera may not be the same model as the first camera and the second camera.
In this embodiment, if the station hall is provided with a plurality of entrance gates and exit gates, a first camera is installed at each entrance gate, and a second camera is installed at each exit gate.
In this embodiment, the height of the standing hall is 2.5 meters, and the inclination angle of the first camera and the second camera is 30 degrees, that is, the included angle between the lens and the horizontal plane is 60 degrees. The tilt angle of the compensation camera mounting is 45 degrees.
In other embodiments, the installation tilt angles of the first camera, the second camera and the compensation camera are other values in the range of 30-60 degrees, depending on the height of the station hall.
In this embodiment, the compensation cameras are installed at corners of the station hall, and shooting ranges of the compensation cameras can cover the whole station hall, so that passengers standing at the corners are prevented from being missed to shoot, and the statistical accuracy is improved.
The undescribed parts of the present invention are the same as or implemented using prior art.
Claims (6)
1. The error control method of the subway station hall people counting system is characterized in that a first camera is installed at a gate entrance of a subway station hall, a second camera is installed at a gate exit, a plurality of compensation cameras are installed in the station hall, the first camera, the second camera and the compensation cameras are all obliquely arranged, the shooting range of the first camera can cover the gate entrance, the shooting range of the second camera can cover the gate exit, and the shooting ranges of the plurality of compensation cameras can cover the whole station hall;
the number of the people who get into the station shot by the first camera in real time is recorded as A1tThe number of the outbound persons shot by the second camera in real time is recorded as A2tThe total number of people in the station hall shot by the plurality of compensation cameras in real time is recorded A3t；
Error delta A of real-time people counting in station halltComprises the following steps:
1) when A3t>At A0,. DELTA.At＝(|A1t-A2t|)×K；
2) When A3tWhen the ratio is less than or equal to A0, delta At＝0；
The models of the first camera and the second camera are the same, and K is the recognition accuracy of pictures or videos shot by the first camera and the second camera; a0 is the set value.
2. The method of claim 1, wherein A0 is any one of 10-30.
3. The method for controlling the error of the subway station hall people counting system according to claim 1 or 2, wherein the first camera, the second camera and the compensation camera have the same or different models.
4. The error control method of the subway station hall people counting system according to claim 3, wherein if the station hall is provided with a plurality of entrance gates and exit gates, each entrance gate is provided with a first camera, and each exit gate is provided with a second camera.
5. The error control method of the subway station hall people counting system according to claim 4, wherein the inclination angle of the installation of the first camera, the second camera and the compensation camera is 30-60 degrees, which is determined by the height of the station hall.
6. The error control method of the subway station hall people counting system according to claim 1, wherein said compensation camera is installed at a corner of the station hall.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910908992.0A CN110855937A (en) | 2019-09-25 | 2019-09-25 | Error control method of subway station hall people counting system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910908992.0A CN110855937A (en) | 2019-09-25 | 2019-09-25 | Error control method of subway station hall people counting system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110855937A true CN110855937A (en) | 2020-02-28 |
Family
ID=69596158
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910908992.0A Pending CN110855937A (en) | 2019-09-25 | 2019-09-25 | Error control method of subway station hall people counting system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110855937A (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017122258A1 (en) * | 2016-01-12 | 2017-07-20 | 株式会社日立国際電気 | Congestion-state-monitoring system |
| CN107871347A (en) * | 2016-09-26 | 2018-04-03 | 河南星云慧通信技术有限公司 | Region demographic method based on video monitoring |
| CN109697435A (en) * | 2018-12-14 | 2019-04-30 | 重庆中科云从科技有限公司 | Stream of people's quantity monitoring method, device, storage medium and equipment |
-
2019
- 2019-09-25 CN CN201910908992.0A patent/CN110855937A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017122258A1 (en) * | 2016-01-12 | 2017-07-20 | 株式会社日立国際電気 | Congestion-state-monitoring system |
| CN107871347A (en) * | 2016-09-26 | 2018-04-03 | 河南星云慧通信技术有限公司 | Region demographic method based on video monitoring |
| CN109697435A (en) * | 2018-12-14 | 2019-04-30 | 重庆中科云从科技有限公司 | Stream of people's quantity monitoring method, device, storage medium and equipment |
Non-Patent Citations (1)
| Title |
|---|
| 范龙飞，姜子政，李海丰，陈新伟: "基于局部密度分类的人数统计算法", 《控制工程》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107688764B (en) | Method and device for detecting vehicle violation | |
| CN104809874B (en) | A kind of traffic incidents detection method and apparatus | |
| CN110675592B (en) | High-altitude parabolic early warning protection system based on target identification and control method | |
| CN100551048C (en) | Multiple-camera supervisory control system and tracking thereof based on the three-dimensional video-frequency dynamic tracking | |
| CN109544946B (en) | Tunnel real-time monitoring management system based on traffic flow big data and implementation method thereof | |
| WO2019136315A3 (en) | Systems and methods for volumetric sizing | |
| TWI466545B (en) | Image capturing device and image monitoring method using the image capturing device | |
| CN111432179A (en) | Intelligent coal conveying belt inspection system and method based on computer vision | |
| CN101699862A (en) | High-resolution region-of-interest image acquisition method of PTZ camera | |
| CN101572804A (en) | Multi-camera intelligent control method and device | |
| CN112489371A (en) | Swimming pool drowning prevention early warning system based on computer vision | |
| US20130194427A1 (en) | Systems methods for camera control using historical or predicted event data | |
| CN107705326A (en) | A kind of intrusion detection method that crosses the border in security sensitive region | |
| CN106558224A (en) | A kind of traffic intelligent monitoring and managing method based on computer vision | |
| JP2018090351A (en) | Elevator system | |
| CN106815575A (en) | Optimization system and method for face detection result set | |
| CN106210553B (en) | Snapshot optimization method and device under shadow shielding | |
| JP2018158842A (en) | Image analyzer and elevator system | |
| CN110855937A (en) | Error control method of subway station hall people counting system | |
| CN110650279A (en) | Design method of camera for monitoring high-altitude object throwing | |
| CN110633648A (en) | Face recognition method and system in natural walking state | |
| CN110759199A (en) | Device and method for detecting electric bicycle trying to enter elevator | |
| TWI537885B (en) | Monitoring method and monitoring system | |
| CN103345653B (en) | Based on the attendance statistical method that multi-cam merges | |
| CN104539875A (en) | Video monitoring method and system |
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 |
Application publication date: 20200228 |
|
| RJ01 | Rejection of invention patent application after publication |