CN105205900A - Dynamic self-adaptive public transport passenger flow statistic device based on video recognition - Google Patents
Dynamic self-adaptive public transport passenger flow statistic device based on video recognition Download PDFInfo
- Publication number
- CN105205900A CN105205900A CN201510697406.4A CN201510697406A CN105205900A CN 105205900 A CN105205900 A CN 105205900A CN 201510697406 A CN201510697406 A CN 201510697406A CN 105205900 A CN105205900 A CN 105205900A
- Authority
- CN
- China
- Prior art keywords
- passenger flow
- people
- processing unit
- central processing
- period
- 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.)
- Granted
Links
Landscapes
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a dynamic self-adaptive public transport passenger flow statistic device based on video recognition. The device is characterized by comprising a binocular camera and a CPU (Central Processing Unit), wherein the binocular camera is arranged above a vehicle door in a perpendicular downward manner and used for carrying out image acquisition on passengers entering the vehicle door; the CPU is in signal connection with the binocular camera and is used for identifying features, acquired by the binocular camera, on an image and obtaining passenger flow statistic results respectively in a passenger flow peak period and a passenger flow sparse period by adopting two algorithms, wherein the feature points comprises double shoulders and the head. The dynamic self-adaptive public transport passenger flow statistic device has the beneficial effects that the transparent capacity peak period is comprehensively considered by combing image recognition algorithms, different image recognition algorithms are applied against different periods, different weighted values are given against different algorithms to accordingly judge the current passenger flow, and therefore the statistic accuracy rates in the passage flow peak period and the sparse period are greatly increased.
Description
Technical field
The present invention relates to a kind of passenger flow statistic device, be specifically related to a kind of dynamic self-adapting bus passenger flow statistic device based on video identification, belong to field of image recognition.
Background technology
In real time, accurate statistics bus passenger flow amount can be public transport company's real-time dynamic scheduling public transit vehicle and provides solid reference.
Current, the device of statistics passenger flow mainly contains two classes: a class is sensor-based passenger flow statistic device, and another kind of is passenger flow statistic device based on image recognition.
Sensor-based passenger flow statistic device, it is the pressure transducer placing some on car pedal, according to the change of pedal pressing force, judges the quantity of trampling people.When the stream of people is sparse, statistics effect is better, but when the crowd is dense, pedal while multiple-people treadled time, just cannot accurately judge stream of people's quantity.
Based on the passenger flow statistic device of image recognition, it is by the identification to the number of people, judges the number of getting on or off the bus, and effectively can avoid the defect that simple sensor designs.The crowd is dense, pedal simultaneously multiple-people treadled time, such passenger flow statistic device has a distinct increment in accuracy than sensor-based passenger flow statistic device, but when the stream of people is sparse, often there is misjudgment phenomenon in the identification of simple number of people feature, such as: easy is the number of people by article identification such as the spherical body of number of people size, toys.
Therefore, how to improve passenger flow statistic device in stream of people's congestion period and the recognition accuracy of the stream of people's sparse period, be a problem needing to solve simultaneously.
Summary of the invention
For solving the deficiencies in the prior art, the object of the present invention is to provide a kind of dynamic self-adapting bus passenger flow statistic device based on video identification, this passenger flow statistic device effectively can improve stream of people's congestion period and the recognition accuracy of the stream of people's sparse period simultaneously.
In order to realize above-mentioned target, the present invention adopts following technical scheme:
Based on a dynamic self-adapting bus passenger flow statistic device for video identification, it is characterized in that, comprising: binocular camera and cpu central processing unit,
Aforementioned binocular camera is arranged in above car door, vertically downward, carries out image acquisition for the real-time passenger to entering car door;
Aforementioned cpu central processing unit is connected with binocular camera signal, for identifying the unique point on the image that binocular camera collects, and adopting two kinds of algorithms to obtain stream of people's statistics of peak traffic period and the stream of people's sparse period respectively, preceding feature point comprises: both shoulders and head;
Aforementioned cpu central processing unit has at least two external interfaces, wherein,
First external interface is connected with door contact interrupter signal, and during car door opening, door contact interrupter disconnects, cpu central processing unit enters time status, and when the duration of car door opening exceedes the threshold value of setting, cpu central processing unit judges that present period is as stream of people peak period, otherwise, be the stream of people's sparse period;
Second external interface is connected with peripheral device signals, sends to peripherals for during passenger flow statistics fructufy by cpu central processing unit.
The aforesaid dynamic self-adapting bus passenger flow statistic device based on video identification, is characterized in that, the algorithm that cpu central processing unit obtains stream of people's statistics of peak traffic period is:
Reduce the weight threshold to both shoulders feature detection, improve the weight threshold to head feature detection, then according to comprehensive weight, the graphics field meeting unique point is counted.
The aforesaid dynamic self-adapting bus passenger flow statistic device based on video identification, is characterized in that, the algorithm that cpu central processing unit obtains stream of people's statistics of the stream of people's sparse period is:
Improve the weight threshold to both shoulders feature detection, reduce the weight threshold to head feature detection, then according to comprehensive weight, the graphics field meeting unique point is counted.
The aforesaid dynamic self-adapting bus passenger flow statistic device based on video identification, it is characterized in that, vehicle opens the door at every turn, and cpu central processing unit all can detect people's stream mode and judge, and using the initial configuration value of this result as video recognition algorithms next time.
Usefulness of the present invention is: combining image recognizer, consider transport power peak period as a whole, for Different periods, use different image recognition algorithms, and provide different weighted value for algorithms of different, with the current passenger flow of this comprehensive descision, substantially increase the statistics accuracy rate of passenger traffic peak period and sparse period.
Accompanying drawing explanation
Fig. 1 is the composition schematic diagram of bus passenger flow statistic device of the present invention;
Fig. 2 is the feature of people in image.
Embodiment
Below in conjunction with the drawings and specific embodiments, concrete introduction is done to the present invention.
With reference to Fig. 1, the dynamic self-adapting bus passenger flow statistic device based on video identification of the present invention, it comprises: binocular camera and cpu central processing unit.
Introduce binocular camera and cpu central processing unit respectively below.
One, binocular camera
Binocular camera is arranged in above car door, and vertically downward, it carries out image acquisition for the real-time passenger to entering car door.
According to the putting position of binocular camera, people is when getting on the bus, and in image, the feature of people is similar to Figure 2, and head is roughly rounded, and both shoulders are roughly rectangle.
Two, cpu central processing unit
Cpu central processing unit is the core of whole passenger flow statistic device, and it is connected with binocular camera signal, for obtaining stream of people's statistics.
Cpu central processing unit obtains the process of stream of people's statistics, is mainly divided into two large steps.
The first step: identify the unique point on the image that binocular camera collects.
Cpu central processing unit to be got on the bus image according to passenger flow, identifies the unique point (comprising: both shoulders and head) with such graphic feature, and counts the graphics field meeting this unique point.Each meets the graphics field of this feature, is all considered to a people.
When carrying out graphic feature identification, cpu central processing unit can carry out feature extraction respectively to both shoulders and head, and calculates comprehensive weight according to predefined weighted value.
Second step: adopt two kinds of algorithms to obtain stream of people's statistics of peak traffic period and the stream of people's sparse period respectively.
In the peak traffic period: cpu central processing unit reduces the weight threshold to both shoulders feature detection, improves the weight threshold to head feature detection, then counts the graphics field meeting unique point according to comprehensive weight.
When passenger traffic peak, the stream of people are crowded time, distance between people and people is very little, it is very near that shoulder leans on, this can produce interference to cpu central processing unit identification shoulder feature, and the recognition accuracy of cpu central processing unit to head feature is higher, therefore, in the peak traffic period, improve the weighted value to the identification of head feature, reduce the identification weighted value to shoulder feature, thus the accuracy rate of passenger traffic peak, stream of people's congestion period people flow rate statistical can be improved.
In the stream of people's sparse period: cpu central processing unit improves the weight threshold to both shoulders feature detection, reduces the weight threshold to head feature detection, then counts the graphics field meeting unique point according to comprehensive weight.
So, in the stream of people's sparse period, to meeting head feature, but be not inconsistent the object of shoulder feature, such as spherical body, can effectively filter out, avoid the mistake identification of the object of the feature to similar head, thus the accuracy rate of the sparse period people flow rate statistical of the stream of people can be improved.
In addition, cpu central processing unit also has at least two external interfaces.The present embodiment only gives two external interfaces, is designated as the first external interface and the second external interface respectively.
1, the first external interface
First external interface is connected with door contact interrupter signal, and during car door opening, door contact interrupter disconnects, and cpu central processing unit enters time status,
(1) when the duration of car door opening exceedes the threshold value of setting, cpu central processing unit judges that present period is as stream of people peak period, will according to predetermined configuration, different weighted values to head and shoulder identification in adjustment image recognition algorithm, the unique point quantity adopting the algorithm of " reduce the weight threshold to both shoulders feature detection, improve weight threshold to head feature detection " to carry out comprehensive statistics to meet head shoulder feature;
(2) when the duration of car door opening does not exceed the threshold value of setting, cpu central processing unit judges that present period is as the stream of people's sparse period, will according to predetermined configuration, the unique point quantity adopting the algorithm of " reduce the weight threshold to both shoulders feature detection, improve weight threshold to head feature detection " to carry out comprehensive statistics to meet head shoulder feature.
Vehicle opens the door at every turn, and cpu central processing unit all can detect people's stream mode and judge, and using the initial configuration value of this result as video recognition algorithms next time.
2, the second external interface
Second external interface is connected with peripherals (such as COM) signal, sends to peripherals for during passenger flow statistics fructufy by cpu central processing unit.
As can be seen here, passenger flow statistic device combining image recognizer of the present invention, consider transport power peak period as a whole, for Different periods, use different image recognition algorithms, and provide different weighted value for algorithms of different, with the current passenger flow of this comprehensive descision, substantially increase the statistics accuracy rate of passenger traffic peak period and sparse period.
It should be noted that, above-described embodiment does not limit the present invention in any form, the technical scheme that the mode that all employings are equal to replacement or equivalent transformation obtains, and all drops in protection scope of the present invention.
Claims (4)
1., based on a dynamic self-adapting bus passenger flow statistic device for video identification, it is characterized in that, comprising: binocular camera and cpu central processing unit,
Described binocular camera is arranged in above car door, vertically downward, carries out image acquisition for the real-time passenger to entering car door;
Described cpu central processing unit is connected with binocular camera signal, for identifying the unique point on the image that binocular camera collects, and adopting two kinds of algorithms to obtain stream of people's statistics of peak traffic period and the stream of people's sparse period respectively, described unique point comprises: both shoulders and head;
Described cpu central processing unit has at least two external interfaces, wherein,
First external interface is connected with door contact interrupter signal, and during car door opening, door contact interrupter disconnects, cpu central processing unit enters time status, and when the duration of car door opening exceedes the threshold value of setting, cpu central processing unit judges that present period is as stream of people peak period, otherwise, be the stream of people's sparse period;
Second external interface is connected with peripheral device signals, sends to peripherals for during passenger flow statistics fructufy by cpu central processing unit.
2. the dynamic self-adapting bus passenger flow statistic device based on video identification according to claim 1, is characterized in that, the algorithm that cpu central processing unit obtains stream of people's statistics of peak traffic period is:
Reduce the weight threshold to both shoulders feature detection, improve the weight threshold to head feature detection, then according to comprehensive weight, the graphics field meeting unique point is counted.
3. the dynamic self-adapting bus passenger flow statistic device based on video identification according to claim 1, is characterized in that, the algorithm that cpu central processing unit obtains stream of people's statistics of the stream of people's sparse period is:
Improve the weight threshold to both shoulders feature detection, reduce the weight threshold to head feature detection, then according to comprehensive weight, the graphics field meeting unique point is counted.
4. the dynamic self-adapting bus passenger flow statistic device based on video identification according to claim 1, it is characterized in that, vehicle opens the door at every turn, and cpu central processing unit all can detect people's stream mode and judge, and using the initial configuration value of this result as video recognition algorithms next time.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510697406.4A CN105205900B (en) | 2015-10-23 | 2015-10-23 | Dynamic self-adapting bus passenger flow statistic device based on video identification |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510697406.4A CN105205900B (en) | 2015-10-23 | 2015-10-23 | Dynamic self-adapting bus passenger flow statistic device based on video identification |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105205900A true CN105205900A (en) | 2015-12-30 |
| CN105205900B CN105205900B (en) | 2017-08-25 |
Family
ID=54953554
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510697406.4A Active CN105205900B (en) | 2015-10-23 | 2015-10-23 | Dynamic self-adapting bus passenger flow statistic device based on video identification |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105205900B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108345820A (en) * | 2017-01-23 | 2018-07-31 | 许继集团有限公司 | High-tension apparatus image-recognizing method and device based on variety of components and component locations |
| CN108446611A (en) * | 2018-03-06 | 2018-08-24 | 深圳市图敏智能视频股份有限公司 | A kind of associated binocular image bus passenger flow computational methods of vehicle door status |
| CN108509914A (en) * | 2018-04-03 | 2018-09-07 | 华录智达科技有限公司 | Bus passenger flow statistical analysis system based on TOF camera and method |
| CN109372344A (en) * | 2018-11-16 | 2019-02-22 | 上海鹿特士环保科技有限公司 | A kind of passenger flow counting Intelligent lock device |
| CN109815787A (en) * | 2018-12-10 | 2019-05-28 | 深圳前海达闼云端智能科技有限公司 | Target identification method, device, storage medium and electronic equipment |
| CN109815936A (en) * | 2019-02-21 | 2019-05-28 | 深圳市商汤科技有限公司 | A kind of target object analysis method and device, computer equipment and storage medium |
| CN111695453A (en) * | 2020-05-27 | 2020-09-22 | 深圳市优必选科技股份有限公司 | Drawing book identification method and device and robot |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2005242646A (en) * | 2004-02-26 | 2005-09-08 | Ntt Data Corp | People flow measuring system, people flow measuring method and people flow measuring program |
| CN101431664A (en) * | 2007-11-06 | 2009-05-13 | 同济大学 | Automatic detection method and system for intensity of passenger flow based on video image |
| CN101714293A (en) * | 2009-12-16 | 2010-05-26 | 上海交通投资信息科技有限公司 | Stereoscopic vision based acquisition method of congestion degree of bus passenger flow |
| CN201717992U (en) * | 2010-05-06 | 2011-01-19 | 杭州惠尔森科技有限公司 | Video passenger detector |
| DE102009046628A1 (en) * | 2009-11-11 | 2011-05-12 | Ifm Electronic Gmbh | Method for detecting covering density in e.g. bus in public passenger traffic, involves determining covering density and/or passenger distribution of transportation region based on depth image |
| CN104021605A (en) * | 2014-04-16 | 2014-09-03 | 湖州朗讯信息科技有限公司 | Real-time statistics system and method for public transport passenger flow |
-
2015
- 2015-10-23 CN CN201510697406.4A patent/CN105205900B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2005242646A (en) * | 2004-02-26 | 2005-09-08 | Ntt Data Corp | People flow measuring system, people flow measuring method and people flow measuring program |
| CN101431664A (en) * | 2007-11-06 | 2009-05-13 | 同济大学 | Automatic detection method and system for intensity of passenger flow based on video image |
| DE102009046628A1 (en) * | 2009-11-11 | 2011-05-12 | Ifm Electronic Gmbh | Method for detecting covering density in e.g. bus in public passenger traffic, involves determining covering density and/or passenger distribution of transportation region based on depth image |
| CN101714293A (en) * | 2009-12-16 | 2010-05-26 | 上海交通投资信息科技有限公司 | Stereoscopic vision based acquisition method of congestion degree of bus passenger flow |
| CN201717992U (en) * | 2010-05-06 | 2011-01-19 | 杭州惠尔森科技有限公司 | Video passenger detector |
| CN104021605A (en) * | 2014-04-16 | 2014-09-03 | 湖州朗讯信息科技有限公司 | Real-time statistics system and method for public transport passenger flow |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108345820A (en) * | 2017-01-23 | 2018-07-31 | 许继集团有限公司 | High-tension apparatus image-recognizing method and device based on variety of components and component locations |
| CN108446611A (en) * | 2018-03-06 | 2018-08-24 | 深圳市图敏智能视频股份有限公司 | A kind of associated binocular image bus passenger flow computational methods of vehicle door status |
| CN108509914A (en) * | 2018-04-03 | 2018-09-07 | 华录智达科技有限公司 | Bus passenger flow statistical analysis system based on TOF camera and method |
| CN109372344A (en) * | 2018-11-16 | 2019-02-22 | 上海鹿特士环保科技有限公司 | A kind of passenger flow counting Intelligent lock device |
| CN109815787A (en) * | 2018-12-10 | 2019-05-28 | 深圳前海达闼云端智能科技有限公司 | Target identification method, device, storage medium and electronic equipment |
| CN109815936A (en) * | 2019-02-21 | 2019-05-28 | 深圳市商汤科技有限公司 | A kind of target object analysis method and device, computer equipment and storage medium |
| CN109815936B (en) * | 2019-02-21 | 2023-08-22 | 深圳市商汤科技有限公司 | Target object analysis method and device, computer equipment and storage medium |
| CN111695453A (en) * | 2020-05-27 | 2020-09-22 | 深圳市优必选科技股份有限公司 | Drawing book identification method and device and robot |
| CN111695453B (en) * | 2020-05-27 | 2024-02-09 | 深圳市优必选科技股份有限公司 | Drawing recognition method and device and robot |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105205900B (en) | 2017-08-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105205900A (en) | Dynamic self-adaptive public transport passenger flow statistic device based on video recognition | |
| CN110458061B (en) | Method for identifying old people falling down and accompanying robot | |
| CN108515990A (en) | A kind of subway carriage carrying situation detection broadcasting system and method | |
| US20160210835A1 (en) | Human body tumbling detection method and device and mobile terminal system | |
| CN104504856A (en) | Fatigue driving detection method based on Kinect and face recognition | |
| CN108509914B (en) | Bus passenger flow statistical analysis system and method based on TOF camera | |
| CN106155312B (en) | Gesture recognition and control method and device | |
| EP2806373A3 (en) | Image processing system and method of improving human face recognition | |
| EP2570963A3 (en) | Systems, devices, and methods for recognizing objects as perceived from a vehicle | |
| CN107248306A (en) | A kind of system for adjusting public transport transport power | |
| CN104732236B (en) | A kind of crowd's abnormal behaviour intelligent detecting method based on layered shaping | |
| CN205263910U (en) | Dynamic self -adapting public transport passenger flow makes statistics of device based on video identification | |
| CN105427582A (en) | Intelligent management method for bus lines | |
| CN104268898A (en) | Method for detecting density of passengers in bus on basis of image analysis | |
| EP3565276A1 (en) | Method for operating a hearing aid and hearing aid | |
| CN104239908A (en) | Intelligent ridership automatic statistical method based on self-adaptive threshold value | |
| CN102592140A (en) | Method for monitoring condition of cyanobacteria bloom | |
| CN107392106B (en) | Human activity endpoint detection method based on double thresholds | |
| CN112036217B (en) | Person state detection method and device based on mouth information | |
| CN108777068A (en) | A kind of traffic flow bottleneck identification method based on multi-dimensions test coil collection period | |
| CN110390313B (en) | Violent action detection method and system | |
| TWI293746B (en) | ||
| CN110929766A (en) | Self-adaptive pedestrian mobile phone attitude identification method based on Gaussian mixture clustering algorithm | |
| CN111079488B (en) | Deep learning-based bus passenger flow detection system and method | |
| CN112863515B (en) | Method for recognizing human tongue and mandible actions by using mobile phone loudspeaker and microphone |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CP03 | Change of name, title or address |
Address after: 14 / F, Hualu building, 717 Huangpu Road, Dalian hi tech Industrial Park, Liaoning 116000 Patentee after: Hualu Zhida Technology Co.,Ltd. Address before: 116023 14 / F, Hualu building, 717 Huangpu Road, Dalian hi tech Industrial Park, Liaoning Province Patentee before: HUALU ZHIDA TECHNOLOGY Co.,Ltd. |
|
| CP03 | Change of name, title or address | ||
| CB03 | Change of inventor or designer information |
Inventor after: Zhang Shiqiang Inventor after: Sun Hongfei Inventor after: Zhao Yan Inventor after: Jia Xiaodan Inventor after: Li Baizhan Inventor before: Zhang Shiqiang Inventor before: Sun Hongfei Inventor before: Li Baizhan Inventor before: Jia Xiaodan |
|
| CB03 | Change of inventor or designer information | ||
| CP01 | Change in the name or title of a patent holder |
Address after: 14 / F, Hualu building, 717 Huangpu Road, Dalian hi tech Industrial Park, Liaoning 116000 Patentee after: Hualu Zhida Technology Co.,Ltd. Address before: 14 / F, Hualu building, 717 Huangpu Road, Dalian hi tech Industrial Park, Liaoning 116000 Patentee before: Hualu Zhida Technology Co.,Ltd. |
|
| CP01 | Change in the name or title of a patent holder |