GLOBAL LICENSE PLATE RECOGNITION SOFTWARE FOR TRAFFIC AND ACCESS CONTROL
CARMEN® ANPRis designed to read all vehicle license plate types in the world at any traffic speed. Toll collection and congestion charging systems, traffic monitoring and security, speed and journey time measurement, bus lane and traffic light enforcement, parking or access control and many other systems can benefit from the fast, exact, automatic identification and recognition capabilities of this ANPR/LPR engine.
CARMEN® ANPR reads license plates from many image sources remarkably fast and with the highest recognition accuracy in its class. It offers country-independent recognition of not only of Latin characters, but also Arabic, Cyrillic, Chinese, Korean, Thai and many more, as well as reflective, non-reflective, personalized and special interest plates that are typical in many U.S. states.
- CARMEN® ANPR offers:
- Industry-leading recognition accuracy (>98% globally)
- Country-independent license plate reading
- Exceptionally fast processing speed
CARMEN® ANPR provides solution for common problems by:
- Handling distorted plate images
- Recognizing also personalized and special interest license plates
- Integrating easily into existing systems through SDK
CARMEN® ANPR is ideal for:
- Free-flow toll collection and traffic management systems
- Section speed control and law enforcement applications
- Parking management and access control
- ITS and security surveillance systems
CARMEN® ANPR’s strength comes from its ability to automatically recognize virtually any license plate of vehicles in moving traffic, regardless of country, state or province, plate or character type, and adapts to environmental changes for 24/7 operation. It is fast and easy with straightforward use with any hardware, compatible with any image source (analog / digital / still images / MJPEG video streams) and supports both Windows and Linux platforms. Its development libraries (DLL, SO) allow flexible integration through their SDK, and in server mode CARMEN® ANPR is able to perform parallel processing and handling of unlimited camera inputs.
GENERAL INFORMATION Purpose Automatic recognition of vehicle license plates – a number plate recognition software for various intelligent traffic systems, security and any access control environments. Supported operating systems Windows (32/64 bit) Linux (32/64 bit) Supported platforms x86_32 | x86_64 | ARMv7 Cortex A8 and above | PPC Minimum system requirements 1 GHz CPU | 512 MB RAM | 1 GB HDD | free port/slot for NNC Licensing One license per application thread, multiple license/controller is available Available neural network controllers USB 2.0 – type A USB – internal, 4 pin PCIe card (x1) mini-PCIe card (x1)(new!) INTERFACE Input Still image from file or memory in any image format (BMP | PNG | JPEG | JPEG2K | RAW) Live analog video input (PAL or NTSC) Live digital camera input Output OCR data License plate number in ASCII/UNICODE text Position of the plate Confidence level in percentage Confidence levels for each characters List of further suggestions for each character Individual result for each plate on an image Color of plate (optional) Country ID (optional) Location of each plate on one image Trigger Can be integrated with any trigger device (recommended when recognizing from live image stream) Software motion detection module is included DEVELOPMENT TOOLS Supported programming languages under Windows C/C++, C# Visual Basic .NET Java Supported programming languages under Linux C/C++, Java In the box Development libraries: .dll, .so files Demo application, sample codes for each programming language Neural network controller Comprehensive digital documentation (also available online) AVAILABLE VERSIONS FreeFlow 5K / 8K / 11K Capacity (images/day) unlimited 5760 / 8640 / 11250 Processing threads 1 / 2 / 4 parallel threads 4 parallel threads Credit buffer unlimited 300 / 240 / 200 Time for 4 new credits (sec) – 60 / 40 / 30
Technical specifications are subject to change without prior notice. This document does not constitute an offer.
- Automatic number plate recognition (ANPR) or automatic license plate recognition (ALPR, LPR) as referred to in the US, is a specialized application in the field of optical character recognition (OCR). OCR, as a technology in general, transforms recorded images of printed characters, numbers and letters to digital text format for post processing, printing, duplication or data management.
HOW DOES AUTOMATIC NUMBER PLATE RECOGNITION WORK?
In practice, car plate number registration happens the following way: pictures or video streams of motor vehicles, which are recorded either by CCTV networks or individual purpose-built IP cameras, are fed into special license plate scanning software that recognizes the plate images and reads the characters, converting them into a digital, processable and storable format.
The ALPR results can then be forwarded to a custom-made user application that utilizes the gathered data for a specific purpose, such as access control, speed enforcement, toll collection, traffic analysis, vehicle identification, etc.
The ANPR process normally takes no more than a few hundredths of a second, so it is possible to register multiple license plates per second, even by using a single camera and also in high-speed environments. Consistently excellent ANPR/LPR performance, however, depends on many technological factors and the ability to adapt to several environmental variables.
The license plate reading software primarily needs a workable image to localize and recognize the vehicle registration plate and to read its characters. In most cases ideal images are provided by systems that employ specifically designed cameras that are equipped with appropriate optics and internal or external lighting – a particularly important issue for surveillance in fast-moving traffic and constantly changing light conditions.
As opposed to cameras installed in controlled internal environments, a 24/7 operating traffic or parking camera should have flexible shutter and brightness controls, to adjust to rapidly changing light conditions, such as direct sun or shadows, pitch black night or blinding headlights. Car plate reading cameras operating in these diverse and constantly changing light conditions should always have built-in high-power infrared (IR) light sources – or they should be connected to an external illuminating unit – with the aim to produce images suitable for automatic license plate recognition.
For a range of our purpose-built recognition cameras, please check our ANPR camera product line.
The resulting ANPR system using a dedicated ANPR camera, with the help of the OCR software is normally able to read the plates of moving vehicles even at 250 km/h (155 mph) in any environment.
Automated number plate recognition may be deployed at any place where vehicle access control, traffic and toll enforcement, license plate number registration and statistical applications are feasible through car plate reading.
As a core technology, ARH’s automatic license plate recognition software is designed to integrate into any application around the world – regardless of the alphanumeric character type, form or color of car number plate or the environmental surroundings.
Our flagship ANPR software: CARMEN® FreeFlow can cover all of these challenges; its strength is in its almost limitless range of utilization, the near 100% rate of recognition and its country-independent software engine.
THE ENTIRE ANPR ROUTINE
To find out more about how CARMEN® fits into the big picture – from triggering through dedicated license plate capture cameras all the way to end-user applications – go to CARMEN.hu.
camera independent • scalable • high speed • reliable • high accuracy • country-independent • customizable • multi-platform • real-time results
Carmen® is the brand name of our cutting edge ANPR software family. It is a flexible and scalable solution to all kinds of traffic, parking, tolling or enforcement applications.
Carmen® ANPR’s strength comes from its ability to automatically recognize virtually any license plate of vehicles in moving traffic, regardless of country, state or province, plate or character type. It is fast and easy, offering straightforward use with any hardware, compatible with various image sources (analog / digital / still images / MJPEG video streams) and supports both Windows and Linux platforms. Its development libraries (DLL, SO) allow flexible integration through their SDK, and in server mode Carmen® ANPR is able to perform parallel processing and handling of unlimited camera inputs.
The CARMEN® ANPR software package consists of the following components:
- Automatic number plate recognition engine (region specific)
- Neural Network Controller (NNC)
- Function libraries
- Demo and test applications
- Tutorial and sample applications both in executable and source code forms
- Licence file
Since image quality is one of the key factors to achieve successful license plate recognition, we recommend using digital cameras made specifically for ANPR purposes. For a range of our purpose-built recognition cameras please check our recognition camera portfolio.
The whole software package comes in the form of an SDK, further integration into any user made business logic is possible through its API, compatible with any system that operates in a Windows or Linux environment.
CARMEN® SOFTWARE LICENSE
To ensure you always have the highest recognition rates covering all the new plate types released, part of our solution is a regular update service for both of your ANPR software and the recognition engine. To find out more about the engine specific licenses, the non-expiration of individual licenses, description of the updating process and the availability of multiple licenses, check out our licensing and update policy page.
In order to accommodate diverse needs, ARH offers different Carmen® versions for different applications:
Carmen® FreeFlow is able to handle massive workloads generated by a multi-lane, multi-camera highway application – potentially processing hundreds of thousands of images per day.
Carmen® 5k / 8k / 11k versions have been designed to manage smaller tasks like less intense roads and parking solutions, these are cost effective versions of the same high standard ANPR/LPR technology.
To help you find the right version for your business needs, please check out our comparative video of Carmen® ANPR versions. Or simply go to the Media tab for more information.
LINKS & DOWNLOADS CARMEN® ANPR SOFTWARE BROCHURE Product description in PDF CARMEN® SOFTWARE VERSIONS A comprehensive review of all CARMEN® software versions ARH PRODUCT CATALOG Comprehensive brochure showcasing all of our traffic, license plate recognition, passport, travel document, ID reader and biometric (AFS/AFIS) products
Use the form below to reach us or send us an email.Contact
NamePlease let us know your name. Email*Please let us know your email address. PhoneInvalid Input SkypeInvalid Input SubjectPlease write a subject for your message. Contact me byInvalid Input *Required information
ANPR/LPR SOFTWARE LIBRARY AND SDK
|The ANPR/LPR software is ideal as an access control engine in all physical environments and even for high speed vehicle traffic.||Two or more linked license plate reading units can provide traffic enforcement systems with accurate average speed measurements, time stamps and plate matching.||Intelligent ANPR/LPR engine can efficiently read the license plate of offending vehicles even at high speed.|
|License plate reading with a fast and accurate ANPR/LPR engine of free flowing traffic provides the most efficient way to verify tolls paid on the road.||Accurate ANPR/LPR reading can provide ITS networks with 24/7 live information about traffic conditions, congestions or accidents.||The accurate ANPR/LPR engine, with its SDK, is designed to provide essential data for traffic analysis and video forensic tasks.|