One of the first South African-designed and manufactured CCTV camera software systems has recently been publically released. The Foresight Security System is designed by Pretoria-based electronic engineer and computer scientist Derek van der Merwe. The system is designed to be a cost-effective way to identify specific threats in real-time and deliver an alarm-video notification to any security company or cell phone anywhere across the globe.
It uses techniques like deep neural networks and can be added to already existing networks with IP cameras. This strategy eliminates the need to re-invest in high-cost cameras that have analytics capabilities. The Foresight Security System was developed in collaboration with software architects and programmers at the Centurion-based Moyo Business Advisory, which specialises in consulting services for change management and data analytics that spans a number of sectors. The security system is currently active at different sites in Gauteng and the Western Cape and is already achieving great results.
van der Merwe wanted to protect his own property and needed greater protection than traditional security measures, such as electric wire fences and electronic beams.
“I set out to design a smart system that could identify certain objects in real-time to put a stop to the endless false alarms I was getting from my beams and cameras every time there were birds, cats or dogs roaming around. “I needed a system that would automatically assess security threats in real time and act fast in hostile situations. Our solution ensures that clients are one step ahead in the event that a security incident or life-threatening situation occurs,” said van der Merwe.
The Foresight Security System was developed by a group of engineers and programmers who specialise in deep learning, AI, neural and machine learning. The system is now available to security companies and specialised installers across the country.
Business Development Director at Moyo Advisory, Dewald Lindeque, has partnered with van der Merwe. Lindeque said that the software behind the CCTV cameras was capable of distinguishing between different items and objects.
“It is able to tell the difference between a dog and a cat, and pretty much any other moveable object out there and only sends out alarms based on the user’s chosen detection preferences and timeframes. In addition, the system has a powerful information distribution capability where the same alarm can be sent to various phones and security companies. This feature can actually realise “crowd sourcing” of security monitoring. Furthermore, the input/output capability, can trigger other hardware to react on a threat, for example switching on a light or a siren. This advanced information management capability, along with advanced information security features, puts the client in control of the protection of personal information,” Lindeque said.
Lindeque noted that the AI will eventually be equipped with facial recognition in order for authorized personnel to move freely through a building or a compound equipped with the security system. The interface, according to Lindeque, is very simple to use so that anyone can be able to protect their property.
“We foresee a vibrant market for this product in farm security protecting farmers and their livestock from raiders 24 hours a day. Although the cameras are live 24 hours a day, they will only send out an alarm when it detects an actual intruder. Our systems will all be backed up by independent power supplies so that they continue to operate even when there is load shedding,” Lindeque added.
Edited by Kojo Essah
Follow Kojo Essah on Twitter
Follow IT News Africa on Twitter