CN107635188A - A kind of video frequency vehicle trace analysis method based on Docker platforms - Google Patents
A kind of video frequency vehicle trace analysis method based on Docker platforms Download PDFInfo
- Publication number
- CN107635188A CN107635188A CN201710803179.8A CN201710803179A CN107635188A CN 107635188 A CN107635188 A CN 107635188A CN 201710803179 A CN201710803179 A CN 201710803179A CN 107635188 A CN107635188 A CN 107635188A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- target vehicle
- video
- characteristic information
- video camera
- 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
Landscapes
- Traffic Control Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention belongs to video frequency vehicle trace analysis technical field, more particularly to a kind of video frequency vehicle trace analysis method based on Docker platforms.Characteristic information of the invention according to target vehicle, traffic direction and the speed of service judge the video camera point position that target vehicle may pass through, according to this video camera point position and the speed of service of the distance between video camera point position and target vehicle thereon, judge that target vehicle reaches the time point of this video camera point position, and based on video flowing realize vehicle running orbit from motion tracking, automatic discrimination and tracking are carried out according to target vehicle traffic direction, manual switching need not be carried out, substantially increase business processing efficiency, and the present invention can carry out vehicle identification and signature analysis to complicated video scene, the accuracy of Tracking Recognition of the present invention is high, good reliability.
Description
Technical field
The invention belongs to video frequency vehicle trace analysis technical field, more particularly to a kind of video car based on Docker platforms
Trace analysis method.
Background technology
With the continuous propulsion that safe city, smart city are built, data also rapid growth is built in camera supervised point position,
Quantity is tens thousand of roads or hundreds thousand of roads, and video includes substantial amounts of people, car, thing information, and information of vehicles is for public security, traffic
It is particularly important etc. for business.
Vehicle tracking analysis is to differentiate a key link violating the regulations, and generally use video is had access to searching and looked into the prior art
Violation vehicle is seen, therefore automatic target following and positioning can not be realized, substantially increases artificial input, business processing efficiency
It is low.
The content of the invention
The present invention in order to overcome the above-mentioned deficiencies of the prior art, there is provided a kind of video frequency vehicle based on Docker platforms with
Track analysis method, the present invention realize Automatic Target Following and positioning, business processing efficiency high.
To achieve the above object, present invention employs following technical measures：
A kind of video frequency vehicle trace analysis method based on Docker platforms, comprises the following steps：
S1, target vehicle is determined, positioning is originally found video camera point position and the time point of target vehicle；
S2, target vehicle tracing task is set, starts to perform tracing task；
S3, extraction are originally found the target vehicle that the video camera of target vehicle is gathered, and identify the feature of target vehicle
Information, traffic direction and the speed of service；
S4, according to the characteristic information, traffic direction and the speed of service of target vehicle judge what target vehicle may pass through
Video camera point position, according to this video camera point position and the speed of service of the distance between video camera point position and target vehicle thereon,
Judge that target vehicle reaches the time point of each video camera point position；
The history video of video camera corresponding to each video camera point position obtained in S5, obtaining step S4, by target vehicle
Characteristic information is compared with the vehicle characteristic information in history video, if vehicle characteristics present in history video are believed
The similarity of breath and the characteristic information of target vehicle is more than setting value, then obtains Vehicle Object, traffic direction in history video
And the speed of service, if the similarity of the characteristic information of vehicle characteristic information present in history video and target vehicle is less than
Setting value, then video camera point position corresponding to the history video is given up；
S6, repeat step S4 and S5, until obtaining the foothold of target vehicle；
S7, the characteristic information for preserving target vehicle；
S8, the driving trace of target vehicle shown on map sequentially in time.
Preferably, the setting value is 90%.
Preferably, step S3 concrete operation step includes：
S31, the moving target being originally found in the video camera of target vehicle is extracted using Background difference；
S32, target vehicle is extracted in moving target；
S33, video capture is carried out to target vehicle；
S34, the characteristic information for establishing deep learning Model Identification target vehicle.
Preferably, the characteristic information by target vehicle in step S5 is compared with the vehicle characteristic information in history video
The concrete operation step of analysis is included：
Each video camera point position obtained in obtaining step S3 corresponds to the history video of video camera, extracts every in history video
One frame picture；
Vehicle detection is carried out successively to each frame picture；
To the detection vehicle detected by vehicle detection according to deep learning model, the characteristic information of recognition detection vehicle；
The characteristic information of the characteristic information and target vehicle that detect vehicle is compared.
Further, type of the characteristic information of the target vehicle including target vehicle, color, the number-plate number, vehicle
Model.
Further, the shooting that the target vehicle tracing task includes the target vehicle tracking time, task is related to is set
Machine monitoring scope and setting value size.
Further, the step S2 operations~step S8 operations are completed on docker platforms.
The beneficial effects of the present invention are：
1), the present invention based on video flowing realize vehicle running orbit from motion tracking, entered according to target vehicle traffic direction
Row automatic discrimination and tracking, without carrying out manual switching, business processing efficiency is substantially increased, and the present invention can be to complexity
Video scene carries out vehicle identification and signature analysis, and the accuracy of Tracking Recognition of the present invention is high, good reliability.
2), the driving trace of target vehicle is shown that exhibition method is directly perceived by the present invention on map sequentially in time, side
Just staff checks.The present invention supports multitask, multichannel video camera to analyze simultaneously, and analysis efficiency is high.
Brief description of the drawings
Fig. 1 is the flow chart of the video frequency vehicle trace analysis method of the present invention；
Fig. 2 is target vehicle identification and the signature analysis flow chart of the present invention；
Fig. 3 is the characteristic information by target vehicle in the step S5 of the present invention and the vehicle characteristic information in history video
The algorithm process flow chart being compared；
Fig. 4 is the middle front end service system of the present invention and the interaction figure of docker platforms.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Site preparation describes, it is clear that described embodiment is only part of the embodiment of the present invention, rather than whole embodiments.It is based on
Embodiment in the present invention, those of ordinary skill in the art are obtained every other under the premise of creative work is not made
Embodiment, belong to the scope of protection of the invention.
As shown in figure 1, a kind of video frequency vehicle trace analysis method based on Docker platforms, comprises the following steps：
S1, target vehicle is determined, positioning is originally found video camera point position and the time point of target vehicle；
S2, the setting target vehicle tracing task are related to camera supervised including target vehicle tracking time, task
Scope and setting value size, start to perform tracing task；
S3, extraction are originally found the target vehicle that the video camera of target vehicle is gathered, and identify the feature of target vehicle
Information, traffic direction and the speed of service；
S4, according to the characteristic information, traffic direction and the speed of service of target vehicle judge what target vehicle may pass through
Video camera point position, according to this video camera point position and the speed of service of the distance between video camera point position and target vehicle thereon,
Judge that target vehicle reaches the time point of each video camera point position；
The history video of video camera corresponding to each video camera point position obtained in S5, obtaining step S4, by target vehicle
Characteristic information is compared with the vehicle characteristic information in history video, if vehicle characteristics present in history video are believed
The similarity of breath and the characteristic information of target vehicle is more than setting value, then obtains Vehicle Object, traffic direction in history video
And the speed of service, if the similarity of the characteristic information of vehicle characteristic information present in history video and target vehicle is less than
Setting value, then video camera point position corresponding to the history video is given up；
S6, repeat step S4 and S5, until obtaining the foothold of target vehicle；
S7, the characteristic information for preserving target vehicle；
S8, the driving trace of target vehicle shown on map sequentially in time.
Specifically, the characteristic information of target vehicle is saved in oracle database, the pictorial information of target vehicle is saved in
In FTP.The characteristic information of association analysis target vehicle and the pictorial information of target vehicle are identified by target vehicle ID, realization is pressed
The space-time for carrying out target vehicle driving trace on map according to time order and function order shows.
As shown in Fig. 2 wherein, extraction is originally found the target vehicle that the video camera of target vehicle is gathered, and identifies mesh
Marking the concrete operation step of the characteristic information of vehicle, traffic direction and the speed of service includes：
S31, the moving target being originally found in the video camera of target vehicle is extracted using Background difference；
S32, target vehicle is extracted in moving target；
S33, video capture is carried out to target vehicle；
S34, the characteristic information for establishing deep learning Model Identification target vehicle.
Specifically, deep learning model is realized using neural network algorithm, sample data is iterated, then logarithm
According to being refined.Intellectuality and the target vehicle that deep learning model is effectively improved algorithm are established using neural network algorithm
Characteristic information identification the degree of accuracy.
As shown in figure 3, the characteristic information by target vehicle in step S5 enters with the vehicle characteristic information in history video
The concrete operation step that row compares analysis includes：
Each video camera point position obtained in obtaining step S3 corresponds to the history video of video camera, extracts every in history video
One frame picture；
Vehicle detection is carried out successively to each frame picture；
To the detection vehicle detected by vehicle detection according to deep learning model, the characteristic information of recognition detection vehicle；
The characteristic information of the characteristic information and target vehicle that detect vehicle is compared.
According to the setting value pre-set, the characteristic information that setting value detection vehicle is more than for similarity stores.
Type of the characteristic information of the target vehicle including target vehicle, color, the number-plate number, vehicle model, vehicle
Brand, car system.
Specifically, the step S2 operations~step S8 operations are completed on docker platforms.
As shown in figure 4, the Docker platforms identify target vehicle ID passes to operation system by http agreements, lead to
The characteristic information of target vehicle and the pictorial information of target vehicle of target vehicle ID mark association analysis are crossed, is realized according to the time
The space-time that sequencing carries out vehicle driving trace on map shows.
Specifically, finally the characteristic information of the target vehicle after association and the pictorial information of target vehicle are according to time order and function
The space-time that order carries out vehicle driving trace on map shows, picture and candid photograph bayonet socket of the content showed including vehicle,
Capture time, type of vehicle, vehicle color, the number-plate number, vehicle brand, car system, vehicle style, car speed, vehicle traveling
Direction.
In summary, the present invention realizes the automatic trace analysis of video frequency vehicle based on Docker platforms, passes through original video
The vehicle characteristic information analyzed in resource is input, according to car speed and traffic direction, platform automatic screening video camera point position,
Go forward side by side driving an aspect ratio pair, according to compare threshold value obtain vehicle travel next passing point, the like tracking target carriage
Driving trace, obtain final foothold.Realize based on video flowing automated vehicle tracking, improve video in target with
Track efficiency.Application scenarios of the present invention can be applied to many aspects of video intelligent track than wide.
Claims (7)
- A kind of 1. video frequency vehicle trace analysis method based on Docker platforms, it is characterised in that comprise the following steps：S1, target vehicle is determined, positioning is originally found video camera point position and the time point of target vehicle；S2, target vehicle tracing task is set, starts to perform tracing task；S3, extraction be originally found the target vehicle that the video camera of target vehicle is gathered, and identify target vehicle characteristic information, Traffic direction and the speed of service；S4, judge according to the characteristic information, traffic direction and the speed of service of target vehicle the shooting that target vehicle may pass through Machine point position, according to this video camera point position and the speed of service of the distance between video camera point position and target vehicle thereon, judge Target vehicle reaches the time point of each video camera point position；The history video of video camera corresponding to each video camera point position obtained in S5, obtaining step S4, by the feature of target vehicle Information is compared with the vehicle characteristic information in history video, if vehicle characteristic information present in history video with The similarity of the characteristic information of target vehicle is more than setting value, then obtain history video in Vehicle Object, traffic direction and The speed of service, if the similarity of the characteristic information of vehicle characteristic information present in history video and target vehicle is less than setting Value, then give up video camera point position corresponding to the history video；S6, repeat step S4 and S5, until obtaining the foothold of target vehicle；S7, the characteristic information for preserving target vehicle；S8, the driving trace of target vehicle shown on map sequentially in time.
- A kind of 2. video frequency vehicle trace analysis method based on Docker platforms as claimed in claim 1, it is characterised in that：Institute Setting value is stated as 90%.
- 3. a kind of video frequency vehicle trace analysis method based on Docker platforms as claimed in claim 1 or 2, its feature exist In step S3 concrete operation step includes：S31, the moving target being originally found in the video camera of target vehicle is extracted using Background difference；S32, target vehicle is extracted in moving target；S33, video capture is carried out to target vehicle；S34, the characteristic information for establishing deep learning Model Identification target vehicle.
- A kind of 4. video frequency vehicle trace analysis method based on Docker platforms as claimed in claim 3, it is characterised in that step The concrete operations that the characteristic information by target vehicle in rapid S5 is compared with the vehicle characteristic information in history video Step includes：Each video camera point position obtained in obtaining step S3 corresponds to the history video of video camera, extracts each frame in history video Picture；Vehicle detection is carried out successively to each frame picture；To the detection vehicle detected by vehicle detection according to deep learning model, the characteristic information of recognition detection vehicle；The characteristic information of the characteristic information and target vehicle that detect vehicle is compared.
- A kind of 5. video frequency vehicle trace analysis method based on Docker platforms as claimed in claim 4, it is characterised in that：Institute State the type of the characteristic information including target vehicle of target vehicle, color, the number-plate number, vehicle model.
- A kind of 6. video frequency vehicle trace analysis method based on Docker platforms as claimed in claim 5, it is characterised in that：If Put the camera supervised scope and setting value that the target vehicle tracing task includes the target vehicle tracking time, task is related to Size.
- A kind of 7. video frequency vehicle trace analysis method based on Docker platforms as claimed in claim 1, it is characterised in that：Institute Step S2 operations~step S8 operations are stated to complete on docker platforms.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710803179.8A CN107635188A (en) | 2017-09-08 | 2017-09-08 | A kind of video frequency vehicle trace analysis method based on Docker platforms |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710803179.8A CN107635188A (en) | 2017-09-08 | 2017-09-08 | A kind of video frequency vehicle trace analysis method based on Docker platforms |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107635188A true CN107635188A (en) | 2018-01-26 |
Family
ID=61100707
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710803179.8A Pending CN107635188A (en) | 2017-09-08 | 2017-09-08 | A kind of video frequency vehicle trace analysis method based on Docker platforms |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107635188A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109766821A (en) * | 2019-01-07 | 2019-05-17 | 深圳英飞拓科技股份有限公司 | Vehicle driving law analytical method, system, computer equipment and storage medium |
| CN110708509A (en) * | 2019-10-10 | 2020-01-17 | 桂林理工大学 | Eye-of-the-sky network video monitoring method for suspect tracking and feature rapid extraction |
| CN113705417A (en) * | 2021-08-23 | 2021-11-26 | 深圳市商汤科技有限公司 | Image processing method and device, electronic equipment and computer readable storage medium |
| CN116958884A (en) * | 2023-09-18 | 2023-10-27 | 杭州靖安防务科技有限公司 | Method and system for target tracking |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1777281A (en) * | 2004-11-17 | 2006-05-24 | 株式会社日立制作所 | Monitoring system using multiple pick-up cameras |
| CN1897015A (en) * | 2006-05-18 | 2007-01-17 | 王海燕 | Method and system for inspecting and tracting vehicle based on machine vision |
| CN102509310A (en) * | 2011-11-18 | 2012-06-20 | 上海电机学院 | Video tracking analysis method and system combined with geographic information |
| CN105632175A (en) * | 2016-01-08 | 2016-06-01 | 上海微锐智能科技有限公司 | Vehicle behavior analysis method and system |
| CN105740855A (en) * | 2016-03-24 | 2016-07-06 | 博康智能信息技术有限公司 | Front and rear license plate detection and recognition method based on deep learning |
| CN106651916A (en) * | 2016-12-29 | 2017-05-10 | 深圳市深网视界科技有限公司 | Target positioning tracking method and device |
| CN106709447A (en) * | 2016-12-21 | 2017-05-24 | 华南理工大学 | Abnormal behavior detection method in video based on target positioning and characteristic fusion |
-
2017
- 2017-09-08 CN CN201710803179.8A patent/CN107635188A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1777281A (en) * | 2004-11-17 | 2006-05-24 | 株式会社日立制作所 | Monitoring system using multiple pick-up cameras |
| CN1897015A (en) * | 2006-05-18 | 2007-01-17 | 王海燕 | Method and system for inspecting and tracting vehicle based on machine vision |
| CN102509310A (en) * | 2011-11-18 | 2012-06-20 | 上海电机学院 | Video tracking analysis method and system combined with geographic information |
| CN105632175A (en) * | 2016-01-08 | 2016-06-01 | 上海微锐智能科技有限公司 | Vehicle behavior analysis method and system |
| CN105740855A (en) * | 2016-03-24 | 2016-07-06 | 博康智能信息技术有限公司 | Front and rear license plate detection and recognition method based on deep learning |
| CN106709447A (en) * | 2016-12-21 | 2017-05-24 | 华南理工大学 | Abnormal behavior detection method in video based on target positioning and characteristic fusion |
| CN106651916A (en) * | 2016-12-29 | 2017-05-10 | 深圳市深网视界科技有限公司 | Target positioning tracking method and device |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109766821A (en) * | 2019-01-07 | 2019-05-17 | 深圳英飞拓科技股份有限公司 | Vehicle driving law analytical method, system, computer equipment and storage medium |
| CN110708509A (en) * | 2019-10-10 | 2020-01-17 | 桂林理工大学 | Eye-of-the-sky network video monitoring method for suspect tracking and feature rapid extraction |
| CN113705417A (en) * | 2021-08-23 | 2021-11-26 | 深圳市商汤科技有限公司 | Image processing method and device, electronic equipment and computer readable storage medium |
| CN116958884A (en) * | 2023-09-18 | 2023-10-27 | 杭州靖安防务科技有限公司 | Method and system for target tracking |
| CN116958884B (en) * | 2023-09-18 | 2024-01-16 | 杭州靖安防务科技有限公司 | Method and system for target tracking |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103069434B (en) | For the method and system of multi-mode video case index | |
| CN102903239B (en) | Method and system for detecting illegal left-and-right steering of vehicle at traffic intersection | |
| CN107635188A (en) | A kind of video frequency vehicle trace analysis method based on Docker platforms | |
| CN101587622B (en) | Forest rocket detecting and identifying method and apparatus based on video image intelligent analysis | |
| CN100545867C (en) | Aerial shooting traffic video frequency vehicle rapid checking method | |
| Morishita et al. | SakuraSensor: Quasi-realtime cherry-lined roads detection through participatory video sensing by cars | |
| CN107085696A (en) | A kind of vehicle location and type identifier method based on bayonet socket image | |
| CN106935035A (en) | Parking offense vehicle real-time detection method based on SSD neutral nets | |
| US20130286198A1 (en) | Method and system for automatically detecting anomalies at a traffic intersection | |
| CN106611162A (en) | Method for real-time detection of road vehicle based on deep learning SSD frame | |
| CN107301776A (en) | Track road conditions processing and dissemination method based on video detection technology | |
| CN111800507A (en) | Traffic monitoring method and traffic monitoring system | |
| KR20200071799A (en) | object recognition and counting method using deep learning artificial intelligence technology | |
| CN106448184A (en) | Identifying method of Vehicles and exit of vehicles | |
| CN106339657A (en) | Straw incineration monitoring method and device based on monitoring video | |
| CN112434566B (en) | Passenger flow statistics method and device, electronic equipment and storage medium | |
| CN106919939B (en) | A kind of traffic signboard tracks and identifies method and system | |
| CN104134067A (en) | Road vehicle monitoring system based on intelligent visual Internet of Things | |
| CN111008574A (en) | Key person track analysis method based on body shape recognition technology | |
| CN113515968A (en) | Method, device, equipment and medium for detecting street abnormal event | |
| CN108965804A (en) | A kind of video structural technology for city security protection | |
| CN114038056A (en) | Skip and squat type ticket evasion behavior identification method | |
| CN113657305B (en) | Video-based intelligent detection method for black smoke vehicle and ringeman blackness level | |
| CN112581495A (en) | Image processing method, device, equipment and storage medium | |
| CN111339834B (en) | Method for identifying vehicle driving direction, computer device and storage medium |
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: 20180126 |