Introduction
Use and Users
Simplified Diagram
Lists
IT Security
Multi-lingual
Contextual Help
Classic Main Screen
Multi-focus screen
Reads and Hits
Preview Page and Popup
Comments
Simple Search
Advanced Search
Data Mining - Statistical
Data Mining - Aggregation
Data Mining - Investigative
Access Management
Hot List Editor
Camera Configuration
Logging
Geographical Information System
Progressive Purging
Hardware Scaling
Introduction
metaBOF is a violation and ANPR data processing software, managing the data from it’s source at the camera, all the way to the visualization on the web interface, giving users efficient access to the collected data.
The metaBOF client is a HTML5 application running in a browser.
The metaBOF software has been designed to allow different services to search through Reads from different types of ANPR cameras (manned, unmanned, fixed, movable, mobile, section control, etc.) in order to use it for various purposes (legal services, investigations, traffic applications, etc.).
Searches can be done based on a particular license plate, but also on time, place, time period, brand of vehicle, country etc., or combinations thereof.
Automatic number plate recognition is a technique using optical character recognition to read license plates on vehicles. The now established term for this technique is the English abbreviation ANPR: Automatic Number Plate Recognition.
In 2008, a first version of metaBOF was created, which, after several revisions and updates, can now be used in entirely different and multiple local areas for both mobile and fixed installations.
Myriade has by now connected cameras from numerous suppliers and keeps adding new ones all the time. A list of these can be made available.
The metaBOF client is a HTML5 application running in a browser.
The metaBOF software has been designed to allow different services to search through Reads from different types of ANPR cameras (manned, unmanned, fixed, movable, mobile, section control, etc.) in order to use it for various purposes (legal services, investigations, traffic applications, etc.).
Searches can be done based on a particular license plate, but also on time, place, time period, brand of vehicle, country etc., or combinations thereof.
Automatic number plate recognition is a technique using optical character recognition to read license plates on vehicles. The now established term for this technique is the English abbreviation ANPR: Automatic Number Plate Recognition.
In 2008, a first version of metaBOF was created, which, after several revisions and updates, can now be used in entirely different and multiple local areas for both mobile and fixed installations.
Myriade has by now connected cameras from numerous suppliers and keeps adding new ones all the time. A list of these can be made available.
Use and Users
metaBOF is used by multiple types of agencies: police, customs, traffic management, environmental agencies and criminal investigators.
ANPR systems can occur in the following configurations: fixed cameras, mobile cameras, section control, truck scales, civil traffic enforcement, SMS Parking, etc.
Owners of ANPR can be Private Services (Parking garage companies, taxis, Towing services, Business parks, etc.), Government Services (Statistics, Congestion charge, Speed controls, Truck locks, Military domains, Customs, Technical Inspection, etc.), Police ( Vehicles, Anonymous operations, Speed controls, criminal investigations, etc.)
A Hot list is a database or list of information. The ANPR system generates an alarm when a license plate is detected that is present in a Hot list
Different types of hot lists can be present: stolen cars, wanted persons, known drug dealers, uninsured cars, emission values of trucks, missed technical examination, etc.
ANPR systems can occur in the following configurations: fixed cameras, mobile cameras, section control, truck scales, civil traffic enforcement, SMS Parking, etc.
Owners of ANPR can be Private Services (Parking garage companies, taxis, Towing services, Business parks, etc.), Government Services (Statistics, Congestion charge, Speed controls, Truck locks, Military domains, Customs, Technical Inspection, etc.), Police ( Vehicles, Anonymous operations, Speed controls, criminal investigations, etc.)
A Hot list is a database or list of information. The ANPR system generates an alarm when a license plate is detected that is present in a Hot list
Different types of hot lists can be present: stolen cars, wanted persons, known drug dealers, uninsured cars, emission values of trucks, missed technical examination, etc.
Simplified Diagram
The Control Center can handle speeding (RADAR or LIDAR), ANPR cameras and Section Control systems.
Lists
- Hot lists or Alarm lists are kept by kept by different organizations and can be:
- Black Lists (plates to find)
- Stolen cars
- Wanted persons
- Known drug dealers
- Uninsured cars
- Emission values of trucks
- Technical Examination
- Known Smugglers
- etc.
- White Lists (plates to overlook temporarily)
- Fire Department
- Police Services
- Disabled people
- Licensees
- etc.
- Fire Department
- Black Lists (plates to find)
- Read Lists contain the reads taken by the cameras connected to the system.
- A read list is the complete list of license plates seen by the ANPR system
- In this list the license plate can be present both readable and scrambled.
- Hit Lists contain the license plates found in both previous lists.
- A hit list is a list of number plates read by the ANPR system that appear in one of the consulted Hot Lists.
- The license plate is in readable format.
IT Security
- Users can access only web applications, running on a separate server
- All data is stored on another machine, all communication is via http(s)
- Access to the Archive Agent (AA) is not possible from outside
- User authentication cookie is passed to ensure rights.
- All search requests are parsed from our custom format to SQL, no SQL injection possible
- ASP.NET Identity Framework is used for user management /access control
- Security is based on roles , which are assigned to groups
- We can add external logon options based on OAUTH
- Every request is checked using an encrypted cookie. The encryption key is in servers configuration file.
- If user should not have access any more, we can force expiration of the cookie from the server side.
- We can implement features like password strength, maximum number of failed logins, account lock-out. Everything comes from the Identity Framework
- SQL DB can be encrypted, access secured by windows or SQL authentication. Only the AA and Indexer have access. Internet user does not talk to those databases.
- PVMyrs can be encrypted, and are only available through AA, never directly.
- All searches and changes in the system are logged
Multi-lingual
metaBOF is fully multilingual. At present, English (EN), Dutch (NL), French (FR), Polish (PL) and Norwegian Bokmål (NO) exist. In a relatively short period languages can be added, preferably in cooperation with the customer.
The basic translations (automated machine translations) already exist for German (GE), Spanish (ES), Italian (IT), Portuguese (PT), Turkish (TR), Bulgarian (BG), Chinese (CN), Russian (RU), Arabic (AR), Slovenian (SL), FI (Finnish) and Swedish (SV) but still require post-processing and contextual adjustments.
Multilingual support is available on all work surfaces: PC, Mac, iPad, and Android devices, as well as related phon
Multilingual support is available on all work surfaces: PC, Mac, iPad, and Android devices, as well as related phon
Contextual Help
Contextual help is available in multiple languages.
Again, the initial version is based on automatic machine translation which needs to be edited and fine-tuned by human readers preferably knowing both the system and the business.
Again, the initial version is based on automatic machine translation which needs to be edited and fine-tuned by human readers preferably knowing both the system and the business.
Classic Main Screen
Viewing, zooming and panning the map, thereby selecting which set of cameras will show Reads in the right hand list.
If no fixed cameras are visible (yellow dots) on the map the list will be empty except whenever a mobile ANPR vehicle happens to drive through the map area while generating reads and the auto-refresh is on.
Hovering over a read in the right hand list shows which camera originated it.
At any time the map will only show the reads listed in the right hand list which is limited to 100 reads (up to 200 reads in Advanced Search) but obviously progressively movable till the end of the data.
If no fixed cameras are visible (yellow dots) on the map the list will be empty except whenever a mobile ANPR vehicle happens to drive through the map area while generating reads and the auto-refresh is on.
Hovering over a read in the right hand list shows which camera originated it.
At any time the map will only show the reads listed in the right hand list which is limited to 100 reads (up to 200 reads in Advanced Search) but obviously progressively movable till the end of the data.
Multi-focus screen
This mode allows for multiple focus areas (one or more cameras, intersections, border passages, etc.) defined by each operator according to his or her needs (morning or evening traffic, day or night traffic, high or low density areas) and share them with each other. Each focus panel has basic search and filtering features. Some panels can communicate with multiple others (preview, map, etc.)
Focus Panel Detail
With black background
Reads and Hits
The right-hand pane shows the last 100 incoming reads from the selected cameras on the map (in this case on the 7 camera locations). When grey these license plates are not searched for.
When [Hits Only] is checked, only Hits are shown and colored according to their priority.
When additionally, notifications are activated hits will produce the following popup form allowing quick on-the-fly verification. Hits are arranged in 3 priorities here: 1: Red, 2: orange, 3: Green. These priorities are applied to each Hot list. There can be more than 3 priorities.
Preview Page and Popup
The General Information Page on tab 1
The detailed information page on tab 2
The Comments page on tab 3
Comments
At the bottom of the folder there is a black field [Comments]. By clicking on it, feedback can be entered.
This allows feedback to be sent to the developers, notes to your self or to link a comment to a read.
Comments can be added to reads, to hits, to Hot lists or directly to the development team or to oneself.
Simple Search
Simple search allows searching on a single criterion out of a selection and includes searching with “wildcard” characters like “?” and “*”.
|
In the combo box next to it, you decide what to look for. Here you can choose which part you wish to search for. The default is License plate ANPR.
You can use "wildcard" characters "?" and "*" to search.
1A?C??? searches all license plates with length 7 of which third, fifth, sixth and seventh characters are variable. 1*BA*13 is not a valid search Following SQL wildcard characters are also supported:
|
The metaBOF system was conceived to fully use the functionality of the browser. This means it will often open new browser windows which the user is expected to close after use.
- When you walk different menus, NO new window will be opened
- If you request a detailed information page, this screen (preview) opens in a separate browser tab.
This is an option to be set by the system administrator. The customer may define this [Search reason] as mandatory input for any search. The search reason can furthermore be forced within a set template such as (for example) XXX######/YYYY where
- XXX can be the following: (dropdown and fill-in):
- JS judicial statement or fine
- TA traffic accident
- TI traffic infraction
- IV intervention
- AP administrative processing
- IR information report
- ###### only numbers
- YYYY year. (also dropdown and fill-in)
Advanced Search
Advanced Search is a multi-criteria search
The possible criteria are the following:
The advanced search further allows:
- Geo Selection (One or more Regions or one or more Streets)
- Geography (zone, region, district, municipality, street within municipality)
- Camera (any camera connected to the system, static or mobile only, lane selection)
- Make Model Color (requires the MMCR Neural Network option)
- Country Code of vehicle (determined through the syntax of the license plate)
- Choice of Hot lists (black or white lists) or selection of license plates
- Date and time continuous or in daily brackets, day or night
- Speed, notwithstanding that ANPR cameras are usually not metrologically calibrated
- Driving Direction
- More sensors as they become available
The advanced search further allows:
- Saving your searches (the criteria) for re-use and sharing with colleagues and to download their result sets (read meta-data).
- Export result sets to CSV, XML (with or without images) and PDF format.
Data Mining - Statistical
These can be organized in several categories from simple to relatively complex. Myriade adds to these continuously based on customer requests or our own insights.
Network usage statistics
Top speed vehicles
Frequent violators
ANPR vehicle route
Network usage statistics
View monthly network usage statistics shows the actual size of the forwarded images and helps to verify the effective use of 3G/4G by comparing this with the incoming invoices.
Show Hits in time is meant to follow a suspected vehicle further after detection by an ANPR camera such as drug dealers moving from one location to another.
Region Statistics is used for traffic management and for administrative management and shows the number of vehicles per region or per camera over a time period.
Top speed vehicles
Top speed vehicles shows maximum speeds recorded by the ANPR cameras. Which are however not metrologically calibrated.
Frequent violators
Frequent violators shows the vehicles in the database with the highest number of violations.
ANPR vehicle route
ANPR vehicle route shows where ANPR police vehicles have patrolled, which helps determining less frequently patrolled locations.
Show Vehicle Trace follows vehicles not via ANPR but via GPS tracker.
Data Mining - Aggregation
These can be organized in several categories from simple to relatively complex. Myriade adds to these continuously based on customer requests or our own insights.
Abandoned vehicles
Cloned license plates
Travel Times
Overtaking detection
Exceeded parking duration
Abandoned vehicles
Abandoned vehicles is used with mobile ANPR vehicles and finds vehicles not moved for a long period of time.
Hostile vehicles looks for vehicles repeatedly found too close to a vehicle (money or prisoner transport, VIP, illegal immigration, etc.).
Cloned license plates
Cloned plates finds identical plates at different locations at impossibly similar moments. One of the vehicles must have a cloned plate.
Travel Times
Travel Times calculates statistically within a time period based on chosen cameras or regions.
Overtaking detection
Overtaking detection is done when the ANPR read sequence changes between two ANPR cameras proving an overtaking maneuver was performed. Could be used for overtaking prohibition by trucks in rainy conditio
Exceeded parking duration
Vehicles with exceeded parking duration is used with mobile ANPR vehicles and finds vehicles parked longer than the maximum allowed time in some zones.
Data Mining - Investigative
These can be organized in several categories from simple to relatively complex. Myriade adds to these continuously based on customer requests or our own insights.
Convoy Analysis
Convoy analysis finds vehicles often found together in a certain area over multiple cameras or regions over a certain time period.
In the table of the results you will find:
- "Candidates" are the Reads (possibly) involved in a convoy. So only the Reads that have been seen close together. -
- "All Reads" are all Reads from the two number plates that were seen on all cameras and within a certain time range. It is greater than or equal to the number of candidates reads. Checking this list indicates whether a license plate is a frequent visitor or was exceptionally visible in that area.
- You can select the "minimum number of times seen together" in the advanced options. But then you have to broaden the scope.
- When you hover a row in the table with the mouse, the row is selected green and the cameras in question become green on the card.
Pattern Matching finds the mathematical cross-section of several sets of read license plates within multiple location and time segments in which we search for the license plate of suspected burglars for example.
Unexpected Vehicles
Unexpected vehicles looks for vehicles seen on a specific date in a specific region but not seen regularly in the past. This requires sufficient reads in the past. This helps finding spotters during the preparation of a burglary.
Arrest on appointment intercepts suspects at the best moments based on their habits and may even catch them red handed. It can search on partial license plate.
Export of Violations
Export of Violations is used for vehicle-free days in streets or neighborhoods and works with white lists. Vehicles rented by customers of hotels in the vehicle-free street or neighborhood should not be fined but are not known on beforehand. All Reads during the vehicle-free day are potential violations (or “candidates”) from which all approved vehicles must now be removed.
This is used for vehicle-free days in a street or neighborhood. We do not work with blacklists but with one or more white lists.
The contents of those white lists are not yet fully known at the time of the Reads. The challenge lies, for example, with vehicles rented by customers of hotels in the vehicle-free street or neighborhood. These are only known later and are allowed to circulate in the area during vehicle-free days.
By definition therefore, all Reads during the vehicle-free day are potential violations (or “candidates”). From that list all approved vehicles must now be removed, and this can only happen a number of days after the vehicle-free day, when the hotels have reporterd the license plates of their guests.
Individual reads can still be removed manually from the generated candidate list. Filling in the reason why is required in this case.
The contents of those white lists are not yet fully known at the time of the Reads. The challenge lies, for example, with vehicles rented by customers of hotels in the vehicle-free street or neighborhood. These are only known later and are allowed to circulate in the area during vehicle-free days.
By definition therefore, all Reads during the vehicle-free day are potential violations (or “candidates”). From that list all approved vehicles must now be removed, and this can only happen a number of days after the vehicle-free day, when the hotels have reporterd the license plates of their guests.
Individual reads can still be removed manually from the generated candidate list. Filling in the reason why is required in this case.
Access Management
This tool allows the management metaBOF users, user groups, organizations, resources (devices, lists and software components), resource unions, permissions, permissions sets, time schedules and view access rights.
This system allows the access management for either a single multi-organization database or for multiple single-organization databases for different organizations, including possible sharing of either resources or data among them all this interfaced through LDAP/SAML to Active Directory system, allowing Single Sign-On (SSO)
A user from a certain organization will be known to metaBOF through LDAP, it will then be part of one (or more) User Groups according to his job attributions, this will give him access to certain Resource Types through Resource Unions which contain the Permission Sets to these Resources, defining what each User can do to or with these Resources (per Type).
- We can organize a customer in multiple organizations with separate LDAP integration. Users can be part of more than one organization. Over LDAP metaBOF interrogates one or more Active Directories (AD) where it gets authentication and some basic info about each user. metaBOF does NOT change this data in any way or form.
- Under that level, we have Users which we can organize in User Groups. Via LDAP their individual identity are defined. Within metaBOF the access control defines what they can do to or with what resources. Users are usually people but can be partnering systems.
- The resources themselves can have different types: devices (cameras and sensors), Lists (Hit list aka alarm lists), applications (software)
- Resources are grouped in Resource Unions. This makes it easier to bundle Permission Sets to a lot of different Resources (devices, lists or software) in one go.
- Permissions define the right to do something to a Resource, per Type. This is defined by the software developers (i.e. .CanRead, .CanConfigure, .CanEdit, .CanDelete, .CanActivate, .CanLog, etc.).
- Permissions are grouped in Permission Sets in order to facilitate the assignment in logical sets of permissions.
- The metaBOF Access Control also has Time Schedules defining periods of time when a certain Permission Set to a certain Resource Union (of a certain Type) allowed to a certain User Group remains valid. Outside the active time period, there is NO access.
- Finally, this complex and potentially sizeable structure is flattened into Access Rights for performance reasons.
- User Retention Period plays a role in all Searches as well. The user can only see search results within his allowed retention period (i.e. 1 hour, a week, 6 months)
A user from a certain organization will be known to metaBOF through LDAP, it will then be part of one (or more) User Groups according to his job attributions, this will give him access to certain Resource Types through Resource Unions which contain the Permission Sets to these Resources, defining what each User can do to or with these Resources (per Type).
Hot List Editor
This tool allow management of Hot lists (aka Alarm lists) both Black lists and White lists of different connection types (BOF, Sync, Dynamic, EXT and manually loaded) both global and local and activating them for cameras or sets of cameras.
Camera Configuration
The user can give more user-friendly names to the cameras.
This usually comes down to a geographical reference.
This usually comes down to a geographical reference.
- street name with in/out
- front/back
This section also can show the camera logs
Clicking on the Camera Name in the Camera Logs will take you to this Camera Properties screen.
A user can, depending upon his or her access rights (ACL) change the alias or the nickname of the camera at will.
Example: “Camera on Main Street at house number 13 direction inner city”
If however the customer wants a specific template to be applied to this alias naming we need to configure this in a configuration file.
Example: all camera aliases need to comply with the following rules:
Clicking on the Camera Name in the Camera Logs will take you to this Camera Properties screen.
A user can, depending upon his or her access rights (ACL) change the alias or the nickname of the camera at will.
Example: “Camera on Main Street at house number 13 direction inner city”
If however the customer wants a specific template to be applied to this alias naming we need to configure this in a configuration file.
Example: all camera aliases need to comply with the following rules:
- Camera number of maximum 5 numbers
- GPS Coordinates
- One of the 4 cardinal points”
- Street name of maximum 10 characters
Logging
Every action or change is logged by metaBOF.
- Thanks to the logging metaBOF is GDPR compliant
- metaBOF can be configured to force users to supply a reason why they perform a certain action or search
- Data Protection Officer (DPO) can therefore trace all actions by any user or even third party systems interfaced with metaBOF.
- metaBOF logs
- Related with Search
- Related with Data Tools
- Related with Exceptions (for developers)
- System Logs
All logs can be sent to ELK stack (ElasticSearch, LogStash, Kibana)
[Add more info and screenshot Kibana]
[Add more info and screenshot Kibana]
Geographical Information System
Myriade uses https://www.openstreetmap.org as Internal Mapping products
Other maps can be used such as Bing Maps, Google Maps, Esri Maps and others
[More info on Leaflet needed + Screenshot + droplet vs bullets]
Other maps can be used such as Bing Maps, Google Maps, Esri Maps and others
[More info on Leaflet needed + Screenshot + droplet vs bullets]
Progressive Purging
Purging of data will happen in 3 steps
- Step 1: Removing the Overview image(s)
- Step 2: Removing all remaining images (no more storage consuming images)
- Step 3: Removing both PVMYR and Database Record (everything)
Note: Only the meta-data is stored in the database. Images are never stored in a database.
Hardware Scaling
Myriade provides an excel sheet to estimate hardware needs (both Application and SQL server) on Virtual Machines for a metaBOF system based on following input:
This tool also allows growth projections over up to 10 years:
- Number of Lanes
- Cars per Day per Lane
- Size of individual PVMYR zip files
- Days the data and images will be kept
- Expected number of concurrent users
This tool also allows growth projections over up to 10 years:
