CN108960327A - A kind of vehicle tracing system and method based on artificial intelligence - Google Patents
A kind of vehicle tracing system and method based on artificial intelligence Download PDFInfo
- Publication number
- CN108960327A CN108960327A CN201810722456.7A CN201810722456A CN108960327A CN 108960327 A CN108960327 A CN 108960327A CN 201810722456 A CN201810722456 A CN 201810722456A CN 108960327 A CN108960327 A CN 108960327A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- subsystem
- tracking
- detection
- artificial intelligence
- 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
- 238000013473 artificial intelligence Methods 0.000 title claims abstract description 16
- 238000000034 method Methods 0.000 title claims description 18
- 238000001514 detection method Methods 0.000 claims abstract description 47
- 238000012216 screening Methods 0.000 claims abstract description 15
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 11
- 238000012549 training Methods 0.000 claims abstract description 11
- 239000000284 extract Substances 0.000 claims description 8
- 238000000605 extraction Methods 0.000 claims description 3
- 238000012806 monitoring device Methods 0.000 claims description 3
- 238000012544 monitoring process Methods 0.000 claims description 3
- 238000012163 sequencing technique Methods 0.000 claims description 3
- 238000007689 inspection Methods 0.000 claims 3
- 230000001537 neural effect Effects 0.000 claims 1
- 206010039203 Road traffic accident Diseases 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- 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
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G06Q50/40—
Abstract
The invention discloses a kind of vehicle tracing systems based on artificial intelligence, the system comprises three subsystems: tracking and managing subsystem, vehicle detection subsystem and vehicle match subsystem, wherein: tracking and managing subsystem is responsible for configuration management, scheduling, coordinates other two subsystems;Vehicle detection subsystem is responsible for detection, is identified and cuts all vehicles in video image;Vehicle match subsystem is responsible for extracting vehicle and compares vehicle pictures feature vector, and calculates similarity, and the highest vehicle of similarity is found.Then the present invention determines vehicle by images match again, improves car tracing efficiency and precision using the vehicle detection subsystem preliminary screening and the consistent vehicle screenshot of vehicle vehicle of convolutional neural networks training.
Description
Technical field
The present invention relates to artificial intelligence computer vision technique fields, and in particular to a kind of vehicle based on artificial intelligence chases after
Track system and method.
Background technique
Nowadays social, vehicle rapidly rises as main traffic and means of transport, ownership, and then bring road
Road traffic problems are especially prominent, and traffic accident also occurs again and again, merely by manpower carry out traffic administration be increasingly difficult to
It meets the requirements.When there is vehicle hit-and-run, simultaneously because environment or artificial origin can not identify the vehicle identification informations such as license plate
When, it is desirable to by manpower, track identification vehicle is practically impossible to realization in a short time.
In computer vision field, convolutional neural networks regard image classification, target detection and positioning, target identification etc.
The effect of feel task is very prominent.The good convolutional neural networks model of one pre-training can be detected quickly and accurately in image
Each object, but guarantee that two necessary conditions of convolutional neural networks precision are a large amount of training sets and efficient calculate
Power.Vehicle picture number can not reach the requirement of training convolutional neural networks.
Summary of the invention
The technical problem to be solved by the present invention is the present invention is in view of the above problems, provide a kind of vehicle based on artificial intelligence
Tracing system and method utilize the vehicle detection subsystem preliminary screening and vehicle vehicle one of convolutional neural networks training
Then the vehicle screenshot of cause determines vehicle by images match again, improves car tracing efficiency and precision.
The technical scheme adopted by the invention is as follows:
A kind of vehicle tracing system based on artificial intelligence, the system comprises three subsystems: tracking and managing subsystem, vehicle
Detect subsystem and vehicle match subsystem, in which:
Tracking and managing subsystem is responsible for configuration management, scheduling, coordinates other two subsystems;
Vehicle detection subsystem is responsible for detection, is identified and cuts all vehicles in video image;
Vehicle match subsystem is responsible for extracting vehicle and compares vehicle pictures feature vector, and calculates similarity, and phase is found
Like the highest vehicle of degree.
The detection of the vehicle detection subsystem uses vehicle detection model, and the vehicle detection model utilizes a large amount of vehicles
Picture is obtained by convolutional neural networks model training.
A kind of car tracing method based on artificial intelligence is led under the management, scheduling, coordination of tracking and managing subsystem
Cross by vehicle picture input vehicle match subsystem extract feature vector, vehicle detection subsystem preliminary screening it is all with start
Then the consistent vehicle pictures of thing vehicle vehicle pass through vehicle match subsystem for the vehicle feature vector of extraction and screening
Vehicle pictures feature vector matched, and calculate similarity, find the highest vehicle of similarity.
The method implementation process is as follows:
1) vehicle picture is chosen, input vehicle match subsystem extracts feature vector and saves in case comparing;
2) tracking and managing subsystem, which is transferred, occurs all crossings real time monitoring video that section is connected to vehicle, and video is defeated
Enter vehicle detection subsystem, the vehicle detection subsystem marks and selects to intercept out all vehicle pictures;
3) vehicle detection subsystem goes out and the consistent vehicle screenshot of vehicle vehicle according to vehicle classification, preliminary screening;
4) the vehicle screenshot after screening is inputted vehicle match subsystem by tracking and managing subsystem, and vehicle match subsystem extracts special
Vector is levied, and calculates the similarity with vehicle feature vector, finds out the highest vehicle screenshot of similarity;
5) vehicle screenshot is navigated to crossing where the monitoring device for transmitting the screenshot by tracking and managing subsystem, then is with the crossing
Starting point continues searching other junction surveillance videos being connected to the crossing;
6) tracking and managing subsystem connects each crossing according to time sequencing, predicts vehicle whereabouts.
The vehicle detection subsystem identifies all types of vehicles by vehicle detection model, passes through the vehicle detection model energy
Enough identify lorry, car, SUV, motor van.
The vehicle detection model is a convolutional neural networks model, by being obtained using the training of a large amount of vehicle pictures.
The invention has the benefit that
The present invention utilizes the vehicle detection subsystem preliminary screening and the consistent vehicle of vehicle vehicle of convolutional neural networks training
Then screenshot determines vehicle by images match again, improves car tracing efficiency and precision.
Detailed description of the invention
Fig. 1 is vehicle tracing system architecture diagram of the present invention.
Specific embodiment
Below according to Figure of description, in conjunction with specific embodiment, the present invention is further described:
Embodiment 1:
As shown in Figure 1, a kind of vehicle tracing system based on artificial intelligence, the system comprises three subsystems: tracking and managing
Subsystem, vehicle detection subsystem and vehicle match subsystem, in which:
Tracking and managing subsystem is responsible for configuration management, scheduling, coordinates other two subsystems;
Vehicle detection subsystem is responsible for detection, is identified and cuts all vehicles in video image;
Vehicle match subsystem is responsible for extracting vehicle and compares vehicle pictures feature vector, and calculates similarity, and phase is found
Like the highest vehicle of degree.
The detection of the vehicle detection subsystem uses vehicle detection model, and the vehicle detection model utilizes a large amount of vehicles
Picture is obtained by convolutional neural networks model training.
Embodiment 2
A kind of car tracing method based on artificial intelligence, under the management, scheduling, coordination of tracking and managing subsystem, pass through by
Vehicle picture inputs vehicle match subsystem and extracts feature vector, and vehicle detection subsystem preliminary screening is all and accident car
Consistent vehicle pictures of vehicle, then by vehicle match subsystem by the vehicle feature vector of extraction and the vehicle of screening
Picture feature vector is matched, and calculates similarity, finds the highest vehicle of similarity.
The method implementation process is as follows:
1) choose vehicle picture (can manually choose), input vehicle match subsystem extract feature vector save in case
It compares;
2) tracking and managing subsystem, which is transferred, occurs all crossings real time monitoring video that section is connected to vehicle, and video is defeated
Enter vehicle detection subsystem, the vehicle detection subsystem marks and selects to intercept out all vehicle pictures;
3) vehicle detection subsystem goes out and the consistent vehicle screenshot of vehicle vehicle according to vehicle classification, preliminary screening;
4) the vehicle screenshot after screening is inputted vehicle match subsystem by tracking and managing subsystem, and vehicle match subsystem extracts special
Vector is levied, and calculates the similarity with vehicle feature vector, finds out the highest vehicle screenshot of similarity;
5) vehicle screenshot is navigated to crossing where the monitoring device for transmitting the screenshot by tracking and managing subsystem, then is with the crossing
Starting point continues searching other junction surveillance videos being connected to the crossing;
6) tracking and managing subsystem connects each crossing according to time sequencing, predicts vehicle whereabouts.
The vehicle detection subsystem identifies all types of vehicles by vehicle detection model, passes through the vehicle detection model energy
Enough identify lorry, car, SUV, motor van.
Embodiment is merely to illustrate the present invention, and not limitation of the present invention, the ordinary skill in relation to technical field
Personnel can also make a variety of changes and modification without departing from the spirit and scope of the present invention, therefore all equivalent
Technical solution also belong to scope of the invention, scope of patent protection of the invention should be defined by the claims.
Claims (6)
1. a kind of vehicle tracing system based on artificial intelligence, which is characterized in that the system comprises three subsystems: tracking pipe
Manage subsystem, vehicle detection subsystem and vehicle match subsystem, in which:
Tracking and managing subsystem is responsible for configuration management, scheduling, coordinates other two subsystems;
Vehicle detection subsystem is responsible for detection, is identified and cuts all vehicles in video image;
Vehicle match subsystem is responsible for extracting vehicle and compares vehicle pictures feature vector, and calculates similarity, and phase is found
Like the highest vehicle of degree.
2. a kind of vehicle tracing system based on artificial intelligence according to claim 1, which is characterized in that the vehicle inspection
The detection for surveying subsystem uses vehicle detection model, and the vehicle detection model passes through convolutional Neural net using a large amount of vehicle pictures
Network model training obtains.
3. a kind of car tracing method based on artificial intelligence, which is characterized in that
Under the management, scheduling, coordination of tracking and managing subsystem, by the way that vehicle picture input vehicle match subsystem is mentioned
Take feature vector, vehicle detection subsystem preliminary screening is all with the consistent vehicle pictures of vehicle vehicle, then passes through vehicle
Matching subsystem matches the vehicle pictures feature vector of the vehicle feature vector of extraction and screening, and calculates phase
Like degree, the highest vehicle of similarity is found.
4. a kind of car tracing method based on artificial intelligence according to claim 3, which is characterized in that the method is real
Existing process is as follows:
1) vehicle picture is chosen, input vehicle match subsystem extracts feature vector and saves in case comparing;
2) tracking and managing subsystem, which is transferred, occurs all crossings real time monitoring video that section is connected to vehicle, and video is defeated
Enter vehicle detection subsystem, the vehicle detection subsystem marks and selects to intercept out all vehicle pictures;
3) vehicle detection subsystem goes out and the consistent vehicle screenshot of vehicle vehicle according to vehicle classification, preliminary screening;
4) the vehicle screenshot after screening is inputted vehicle match subsystem by tracking and managing subsystem, and vehicle match subsystem extracts special
Vector is levied, and calculates the similarity with vehicle feature vector, finds out the highest vehicle screenshot of similarity;
5) vehicle screenshot is navigated to crossing where the monitoring device for transmitting the screenshot by tracking and managing subsystem, then is with the crossing
Starting point continues searching other junction surveillance videos being connected to the crossing;
6) tracking and managing subsystem connects each crossing according to time sequencing, predicts vehicle whereabouts.
5. a kind of car tracing method based on artificial intelligence according to claim 4, which is characterized in that the vehicle inspection
It surveys subsystem and all types of vehicles is identified by vehicle detection model, lorry, small sedan-chair can recognize that by the vehicle detection model
Vehicle, SUV, motor van.
6. a kind of car tracing method based on artificial intelligence according to claim 5, which is characterized in that the vehicle inspection
Survey model is a convolutional neural networks model, by being obtained using the training of a large amount of vehicle pictures.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810722456.7A CN108960327A (en) | 2018-07-04 | 2018-07-04 | A kind of vehicle tracing system and method based on artificial intelligence |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810722456.7A CN108960327A (en) | 2018-07-04 | 2018-07-04 | A kind of vehicle tracing system and method based on artificial intelligence |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108960327A true CN108960327A (en) | 2018-12-07 |
Family
ID=64485449
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810722456.7A Pending CN108960327A (en) | 2018-07-04 | 2018-07-04 | A kind of vehicle tracing system and method based on artificial intelligence |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108960327A (en) |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103248867A (en) * | 2012-08-20 | 2013-08-14 | 苏州大学 | Surveillance method of intelligent video surveillance system based on multi-camera data fusion |
CN104699726A (en) * | 2013-12-18 | 2015-06-10 | 杭州海康威视数字技术股份有限公司 | Vehicle image retrieval method and device for traffic block port |
CN105893510A (en) * | 2016-03-30 | 2016-08-24 | 北京格灵深瞳信息技术有限公司 | Video structurization system and target search method thereof |
WO2017117359A1 (en) * | 2015-12-30 | 2017-07-06 | 3M Innovative Properties Company | Automatic learning for vehicle classification |
CN107808512A (en) * | 2017-09-08 | 2018-03-16 | 北京悦畅科技有限公司 | Car tracing method and device |
-
2018
- 2018-07-04 CN CN201810722456.7A patent/CN108960327A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103248867A (en) * | 2012-08-20 | 2013-08-14 | 苏州大学 | Surveillance method of intelligent video surveillance system based on multi-camera data fusion |
CN104699726A (en) * | 2013-12-18 | 2015-06-10 | 杭州海康威视数字技术股份有限公司 | Vehicle image retrieval method and device for traffic block port |
WO2017117359A1 (en) * | 2015-12-30 | 2017-07-06 | 3M Innovative Properties Company | Automatic learning for vehicle classification |
CN105893510A (en) * | 2016-03-30 | 2016-08-24 | 北京格灵深瞳信息技术有限公司 | Video structurization system and target search method thereof |
CN107808512A (en) * | 2017-09-08 | 2018-03-16 | 北京悦畅科技有限公司 | Car tracing method and device |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US9436877B2 (en) | Pedestrian right of way monitoring and reporting system and method | |
US8885929B2 (en) | Abnormal behavior detection system and method using automatic classification of multiple features | |
CN105702048B (en) | Highway front truck illegal road occupation identifying system based on automobile data recorder and method | |
US9704201B2 (en) | Method and system for detecting uninsured motor vehicles | |
Haloi et al. | A robust lane detection and departure warning system | |
Ke et al. | Motion-vector clustering for traffic speed detection from UAV video | |
CN103294775A (en) | Police service cloud image recognition vehicle management and control system based on geographic space-time constraint | |
CN108877269A (en) | A kind of detection of intersection vehicle-state and V2X broadcasting method | |
CN110634306A (en) | Method and device for determining vehicle position, storage medium and computing equipment | |
Feng et al. | Mixed road user trajectory extraction from moving aerial videos based on convolution neural network detection | |
CN104065920A (en) | Vehicle monitoring and tracking method, system and server | |
Singh et al. | IoT based weapons detection system for surveillance and security using YOLOV4 | |
Fernández et al. | Robust real-time traffic surveillance with deep learning | |
Gasparini et al. | Anomaly detection, localization and classification for railway inspection | |
Kataev et al. | Method to estimate pedestrian traffic using convolutional neural network | |
CN113870551B (en) | Road side monitoring system capable of identifying dangerous and non-dangerous driving behaviors | |
Balci et al. | Front-View Vehicle Damage Detection using Roadway Surveillance Camera Images. | |
CN108960327A (en) | A kind of vehicle tracing system and method based on artificial intelligence | |
Xu et al. | A novel method for people and vehicle classification based on Hough line feature | |
Kumar et al. | Smart City Traffic Management and Surveillance System for Indian Scenario | |
WO2020210960A1 (en) | Method and system for reconstructing digital panorama of traffic route | |
CN105206060A (en) | Vehicle type recognition device and method based on SIFT characteristics | |
Joshy et al. | Improved detection and tracking of lane marking using Hough transform | |
Purica et al. | A railroad detection algorithm for infrastructure surveillance using enduring airborne systems | |
Gupta et al. | Military surveillance with deep convolutional neural network |
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: 20181207 |