20 February 2012
How machine vision technology brings relief to Dubai commuters.
Dubai's rapid economic growth in the past 10 years has transformed the city. With a population increasing from 860,000 in 2000 to 1.9 million* in 2011 and the number of registered vehicles on the road reaching just over 1 million* traffic chaos and parking have become a daily headache.
How do you solve a problem like traffic congestion?
The United Arab Emirates and the government of Dubai have tackled traffic and parking congestion issues by implementing several measures over the past few years. New infrastructure plans include a AED 28 billion (US$7.5bn) metro system to be completed in 2012, which has provided some relief to commuters totalling 6.8 million trips in its first year and 38.8 million trips with the addition of several lines in 2010. Taxis with 81.8* million trips in 2010 however remain the most commonly used method of public transportation in Dubai adding to the traffic woes. The authorities have also funded a AED 3 billion (US$865m) annual road building programme as well as the implementation of an Intelligent Transportation Systems (ITS) project with a purpose-built traffic management centre at its heart, the introduction of tolling systems on major roads into the city and a complete review of parking facilities and regulations to optimize traffic flow and related issues in congested areas.
The end of parking woes?
With over 1 million vehicles on the road, demand for multi-storey car parks has shot up, especially in the Central Business Districts in Deira and Bur Dubai. In 2010 alone the Dubai Roads and Transport Authority (RTA) announced a plan to build at least 50 multi-storey parking plazas in Dubai to cope with the increasing parking problems. Each new parking facility is required to be equipped with CCTV and a centralised and automated payment system.
HNAPR is an ANPR solution developed by Dubai based Niaar ITS LLC. HNAPR is a rugged and fully automated machine vision based system that relies on a digital industrial camera from Allied Vision technologies that is used for plate recognition in check-point gates for parking management or plate replacement centers applications. The system is built to perform 24/7 in temperatures up to 50°C and is ruggedized to sustain dust and strong sun light.
Niaar’s HNAPR is designed to simplify and speed up the management of all vehicle entry and exit transactions inside a parking area in order to optimize gate management and limit waiting time, thus reducing street traffic congestion. As vehicles go through the gate the system captures burst images of the license plate, which are then analyzed by the system’s software and stored into a database alongside other parametres such as time and date for future referencing. A printer automatically issues receipts to users. Stored image data and parametres help improve the facility’s overall security level and provide vital statistics allowing facility managers to monitor parking usage, analyze traffic levels or even restrict parking usage to certain vehicles based on the date or time of day.
The technology behind parking automation
The HNAPR system consists of a Prosilica GC1290 camera from Allied Vision Technologies fitted with a high quality OEM lens and mounted inside a dust proof IP65 enclosure on top of a 1 meter tall post at a slight angle approximately 2 to 3 metres away from the “target” , i.e. a vehicle’s license plate and can also operate with maximum accuracy from a distance up to 15 metres away from its target. The system is paired with an IR flash for night-time imaging. The Prosilica GC1290 is a one Megapixel digital camera with a Gigabit Ethernet interface (GigE Vision®) and features the Sony ICX445 ExView HAD sensor that provides excellent sensitivity and near IR response.
The camera is connected to an industrial PC running Windows across an Ethernet network via Cat6 Ethernet cable or fiber optics depending on the distance. A server unit is located at each gate entrance. Each license plate character is defined in the software and recognized via an OCR process. The Niaar ANPR software is programmed using C++. The system is installed under 2 hours and is low maintenance, requiring re-calibration only every 6 months to maintain 95% + accuracy. Other settings such as shutter speed, aperture, gain and binning can be easily controlled and adjusted remotely via the software to adapt to local lighting conditions.
New camera technology for parking automation and other ITS applications
A new generation of machine vision cameras specially designed for outdoor applications such as ITS are now available on the market. The Prosilica GT from Allied Vision Technologies offer an array of features that allows ITS integrators to improve the performance and reliability of future traffic solutions. The Prosilica GT cameras feature the latest Sony and Kodak sensors that combine high resolution and fast frame rates to capture multiple shots of the same license plate for added ANPR accuracy, configurable I/O to synchronize the image capture process with traffic system peripherals, low-latency trigger for timely image capture, IEEE1588 Precision Time Protocol to manage clock synchronization of multiple devices such as sensors and lights across an Ethernet network and motorized lens control to adjust to varying lighting conditions.
* Source: Government of Dubai / Dubai Statistics Centre