printer friendly version

LPR is an integrated solution

Licence plate recognition (LPR) is no longer a new or ground-breaking technology and we see it deployed everywhere, from city streets to business parks and residential estates. The value of LPR, however, is far more than simply taking a picture of a licence plate.

The power of LPR is only experienced when it is integrated with other technologies, from video analytics and AI, through to connecting to external databases of stolen or suspect vehicles (or internal white or blacklists).

Monique East from OneSpace says OneSpace includes LPR in its services. “We have recently added high-end cameras to our portfolio which are able to do significantly more than only read the plate. We can now also determine the make, model and colour of the vehicle. This has changed the way we are able to pre-emptively act when community policing is necessary in a given SNIPR precinct.”

Nick Grange from XtraVision, distributors of FF Group LPR solutions and Hanwha cameras, looks at what can be done when the right integrated solution is in place.

Parking solutions: With automated access control to parking in parking spaces or logistic centres, the system can show the way to a particular parking area, alert on improper parking and manage individual tenants’ parking. The system detects number plates, provides access control and can allow entry to a particular level.

Safe and smart cities: The technology used in this environment can provide real-time data and alerts for criminal and civil offence prevention, traffic control and alerts on traffic rules violations.

Integrated access control systems: Systems can provide traffic control in the area of a warehouse, alerting to traffic accidents, wrong-way driving and more.

Do we still need LPR-specific cameras?

With the level of AI some companies claim to have developed, is it still necessary for companies to pay extra for a camera designed specifically for LPR. With all the buzz around AI, one would think any camera can be used and the AI analytics will interpret the licence plate.

The two work hand in glove, says East. “LPR cameras will still have their place in the face of increasing intelligence. AI and video analytics are only as powerful as their ability to be linked to the correct information and the same can be said of licence plate recognition. The true power to effect change lies in gaining information in context.

“Take SNIPR for instance. The high-end cameras we mentioned are underpinned by our ability to link into various law enforcement databases. Essentially, we have the technology to capture an image, analyse, decrypt and interpret it and then inform immediate actions to the relevant authorities based on vehicles which have already been linked to suspicious activity.

“In other words, our ability to process data into intelligent information is directly linked to our ability to read the plates. These things should ideally work in concert with each other.”

Grange says successful LPR is also dependant on the following requirements (when using Arteco or FF Group LPR solutions):

• Licence plate is more than 130 px in width.

• Licence plate is readable.

• Vertical angle is less than 30°.

• Horizontal angle is no more 30°.

• Tilt angle is less than 5°.

Solutions available

Hi-Tech Security Solutions asked OneSpace and XtraVision to describe the solutions they provide, which include or are part of a general LPR system. We also approached DeepAlert for some information on its AI software and how it can be used to support the LPR function.

DeepAlert’s Mark Smuts explains that the DeepAlert AI video analytics system can detect different types of vehicles (car, minibus, commercial vehicle, bus and motorcycle) and this can be used as a rule to trigger LPR. For example, certain lanes on freeways are designated for public transport only (bus, minibus) during certain hours.

“The DeepAlert software can detect vehicles other than those designated in those specific lanes during those specific hours and trigger the LPR cameras to capture the licence plate of the offending vehicle. As DeepAlert can be connected to practically any ONVIF-compliant camera, all that is required is the DeepAlert Export API, which enables connection to the LPR interface. The same LPR camera can be used for analytic detection. The service is sold using the SaaS model with a monthly licence fee per camera connected to the system.”

Grange says XtraVision offers the Wisenet range, combined with ZKTeco, FF Group software and/or the Arteco VEMS. The company can provide a single operator platform to fit a host of applications.

OneSpace Access Control has the ability to provide business estate or community solutions in part, or end to end. “Our solution provides seamless integration of previous silos of information within the business – from opening the gate, linking entry control to guest and vehicle information, managing contractor permissions to site, capturing and running licence plates for cloned or suspicious vehicle activity and alerting necessary authorities,” says East.

“OneSpace has even established intelligent methods of managing targeted communication to members of the community. As all of these are cloud services, this means real-time reporting on all these instances. At OneSpace, we are looking to the future of integrated technology and are excited to be pioneering in this field where we can make a difference to South African communities.”

For more information contact:

• DeepAlert, +27 82 465 9590, [email protected], www.deepalert.ai

• OneSpace, +27 31 035 0941, [email protected], www.one-space.co.za

Share this article:

Further reading: