Automatic number-plate-recognition

1. BY
Devang Tailor

2.  Automatic Number Plate Recognition (or 'number plate recognition')
is a special form of optical character recognition (OCR).
 It enables computer systems to read automatically the registration
number (license number) of vehicles from digital pictures.
 Reading automatically the registration number means transforming
the pixels of the digital image into the ASCII text of the number plate.
 OCR technology scans paper documents and turn them into
electronic, editable files.
 Capturing of fast moving vehicles needs special technique to avoid
motion blur . To ensure the right image quality short shutter time is
used with high-power illumination.

3.  The best illumination is the IR, because the retro-reflective plates
reflect this kind of light and it is undetectable for the human eye.
 This combination works fine during day and night and provides
constant good image quality.
 Only dedicated ANPR cameras meet these requirements, which
provide flexible shutter control with built-in IR flash and able to
catch the vehicles up to 250km/h which is suitable for all kind of
license plate reading applications.

4. √ Traffic Law Enforcement. Detection and prosecution for:
√ running red lights
√ driving at a point-in-time speed over the speed limit
√ Traffic Law Enforcement. Detection and interception of:
√ Unregistered Vehicles
? Driving at an average speed over the speed limit
??Vehicles owned by currently Unlicensed Drivers
√ Public Safety. Deterrence of unsafe practices, e.g.
√ running red lights, speeding
? driving unregistered vehicles
??driving while unlicensed

5.  Applications:
 Intelligent Transportation Systems (ITS),
 security systems
 motorway toll collection (highway toll collection),
 traffic analysis,
 police law enforcement,
 state border control,
 enforcement of traffic regulations,
 vehicle theft prevention,
 automation and simplification of airport,
 harbor and warehouse logistics,
 security monitoring of roads and checkpoints
 vehicle surveillance

6.  ANPR has two technological issues:
 The quality of license plate recognition software with its
applied recognition algorithms, and
 The quality of image acquisition technology, the camera and
the illumination.
 The very key factor is the number plate recognition software.
 A sophistication recognition software should have following
features
 the highest recognition accuracy
 the fastest processing speed
 the most type of number plates it can handle,
 the widest range of picture quality it can handle,
 the most tolerant against distortions of input data .

7.  The key technological parts of an APNR system:
 high accuracy Optical Character Recognition (OCR) – is an
essential requirement
 technology image acquisition technology - determines the
average image quality the number plate recognition algorithm
has to work on.

8. Camera
& OCR
Tightly-
Coupled
Processing
Police
Cars
Alerts

9. Camera
& OCR
Tightly-
Coupled
Processing
Sources of
Data-Sets
Police
Cars
Alerts

10. Camera
& OCR
Tightly-
Coupled
Processing
Sources of
Data-Sets
Operational
Policing
Police
Cars
Alerts
Alerts
OnlyAlerts

11.  Alliances of purveyors and purchasers suggest that
data extraction is accurate and reliable ... BUT ...
 Very little evidence is publicly available
 There appear to be no independent tests
 Many factors reduce reliability, including:
 the nature and condition of the registration plates
 the condition of the camera lens
 the conditions of the light-path and back-lighting
 The extraction is by its nature 'fuzzy',
and confidence threshholds have to be set
 Reliable extraction of the registration data may be
as low as 70% even under favourable conditions

12.  Detection and Interception of:
 Wanted Vehicles, in particular:
 'Reported Stolen'
 'Get-Away Cars'
 Vehicles associated with Persons of Interest
 Dependent on:
 Real-Time Acccess to ...
 Real-Time-Maintained Data Sources

13.  Retrospective Analysis of Vehicle Movements:
 Detection of Duplicates
 False Registration Numbers
 Retrospective Inferences about
Owner Location and Movements
 Retrospective Inferences about
Co-Location, and Co-Location Frequency, of:
 Vehicles
 People
 Real-Time Inferences about Location, Co-Location

14. Camera
& OCR
Operational
Policing
Police
Cars
Alerts
All
Sightings

15. Camera
& OCR
Central
Processing
& Storage
Operational
Policing
Police
Cars
Alerts
All
Sightings
All
Sightings

16.  Different CARMEN software versions are used for different
ANPR application areas
 There are basically 3 versions
 CARMEN Free Flow : general-purpose license plate recognition
engine , used for processing images of fast moving vehicles
 CARMENParking Lane: developed for parking management
applications where automatic identification of vehicles is
needed.
 CARMEN Parking: cost-effective solution, which is ideal for
parking, vehicle entry and access control systems. It is capable
of the automatic detection, identification and recognition of the
registration number of any vehicle.

17.  It is highly accurate, fail safe and stable, fast and intelligent.
 it provides the highest technological performance
 It can be integrated into practically any ANPR system
 Features:
 offers general-purpose license plate reading from digital
pictures of any type of sources.
 provide high quality car plate recognition for very different
type of images and plates with default settings.
 98.5% recognition rate (correct reading/total number of input)

18.  specially designed for parking systems, security systems and
access control systems.
 It has a unique integration level of license plate recognition
technology, the video acquisition technology and the processing of
multiple images of video streams.
 It comes with an industrial quality native driver for image
acquisition and video streaming, providing a plug-and-play core
number plate recognition module (software and capture card) for
car park and vehicle access control systems.

19.  Features:
 offers flexible,
 low maintenance,
 effective and intelligent technology for car park operation.
 reduces costs as well as increasing efficiency and security.
 the Carmen Parking ANPR package contains the following
elements:
 Automatic number plate recognition engine
 FXVD4mc_s grabber card
 Functions library (ANSI C/C++, MSVC/C++, MSVB and
Delphi, ActiveX, OCX)
 Demo and test applications
 Tutorial and sample programs both in executable and source
code.
 High Resolution Camera equipped with a special and
synchronized IR flash.(optional)

20.  A parking lot equipped with Carmen Parking can provide:
 Flexible and automatic vehicle entry to and exit from a car park
 Management information about car park usage
 Improved security for both car park operators and car park users
 Improved traffic flow during peak periods
 Vehicle recognition through date and time stamping as well as
exact location
 Comprehensive database of traffic movement can be managed
 Applications:
 car park management,
 car parking automation,
 occupation management,
 automation of access control,
 security access control,
 flotilla management (car pool management),
 automation of vehicle entry to and exit from car parks,
 inventory management,
 state border control

21.  used for one or two lanes of access control systems.
 With integration CARMEN Parking Lane into parking
management systems, it is possible to automate vehicle entry to
and exit from a car park or a secure zone and to use the
recognized registration number for inventory management.
 specially developed for parking management applications where
automatic identification of vehicles is needed.
 provides accurate and quick identification of vehicles without
human assistance.

22.  ANPR consists of following hardware products :
 Image capturing -FXCAM LPR/ANPR camera series : to provide
good quality images
 ARH Car PC for Mobile ANPR applications : is a mobile PC with
unique features like GPS receiver and GPS antenna.
 Video signal digitizing -FXVD4 frame grabber card : the primary
function of this card is digitizing the video signal. FXVD4mc_s is a
high quality 4-channel frame grabber card

23.  The smart shutter technology enables you to make an adaptive
image capturing system, controlled automatically by software.
 Effective range is up to 20 meters (65 ft) but it may vary
depending on the specific environmental conditions.
 Ideal for both moving traffic and access control (parking
management) applications.
 Image quality:
 Resolution: 752X480 pixels, ROI support
 Image processing speed: 60 frames/sec at full image size
 Continuous, or triggered buffering for 60 full size images
 Lossless RAW image format
 Adjustable gain and shutter time 1/100 to 1/30000 sec
 No extra illumination required

24.  Lens:
 High quality motorized zoom and iris (effective range 3-20m)
 IR pass filter included
 High speed and ultra sensible image sensor :
 CMOS monochrome WideVGA (pixel size 6ľm x 6ľm)
 Global Shutter TrueSNAP
 Dynamic range up to 100dB
 High performing illumination :
 High power IR LEDs - ideal for ANPR day & night
 Strong sunlight / shadow correction
 Power consumption :
 Isolated 24-28V AC or 30-36V DC input (50-60Hz)
 Optional 12V DC (FXCAM202)
 Less than 15W (26V AC) [depending on heating ON/OFF]
 Reverse polarity protection

26.  FXVD4mc_s is a high quality 4-channel frame grabber card
 function of this card is digitizing the video signal by an Bt878
video capture chip.
 To improve the reliability and stability of a stand-alone system the
FXVD4mc_s has a secondary function the watchdog.
 This function helps to reboot the computer when a system crashes
 On the card there is a RISC co-processor which serve the hardware
protection of the software

27.  Interface : PCI 2.1 AD
 conversation : Bt87
 Input : 4 x BNC video, 75 Ohm
 Watchdog : Programmable with 1-250 sec. delays
 Supported Operating System :
 Windows Vista 32bit, XP, 2003, 2000
 Linux
 Programming Languages :
 Visual C/C++
 Visual Basic
 Borland Delphi
 Visual Basic.NET
 C#
 Java
 GCC and Java under Linux

28.  Additional Tools :
 SDK for easy integration
 DLLs and ActiveX components
 Video Formats : PAL or NTSC (75 W, 1 Vp-p)
 Video Framing : Full, Field, Half @ any / odd / even
 Image Output : Memory, file
 File Types : BMP, JPEG
 File Formats : Grayscale, RGB15, RGB16, RGB24, RGB32,
YUV422
 Documentation :Reference Manual in electronic format
 System Requirements : Free PCI slot with Bust Mastering
support
 Power consumption : approx. 450mA
 Size : 97 mm x 120 mm

30.  PC-based system which is specially designed for mobile
installation
 The anti-vibration protection and the anti-shock protection makes
its usage reliable in vehicles.
 it has unique features like
 Intelligent (programmable) UPS
 GPS receiver
 GPS antenna
 Advantages:
 Maintenance free operation
 No moving parts
 Configurable

31.  Processor : Intel Core Duo 1.2 GHz
 RAM : 1 GB DDR2
 Graphics : Integrated Intel graphics controller
 HDD : Automotive 40GB (Flashdrive: optional)
 Network : Gigabit LAN controller
 Input/output:
 2 x RJ45 LAN
 2x serial
 3x USB2.0
 DVI + VGA out
 PS2 for keyboard
 Mic + line in, line out
 Composite + S-video TV out
 Power: Input voltage: 9-30 VDC (5.4A @ 9DC)
 Power consumption: max. 50W (typical 15-20W)
 Power Module with UPS :
 Smart Battery, Li-Polymer, 3000mAH
 GPS receiver and antenna included
 Windows XP Professional included

34.  Parking
 Access Control
 Motorway Road Tolling
 Border Control
 Journey Time Measurement
 Law Enforcement

35. THANK YOU