Video analytics has revolutionised how businesses and organisations approach security. The global market is expected to grow over the next decade as new technological developments allow for real-time observation, detection and reporting of actions within a monitored area. Artificial intelligence (AI) has also spurred this revolution, enabling businesses to derive actionable insights from data gathered in the form of video.
However, while there is a lot to be excited about when it comes to the newfound capabilities of video technology, especially as more enterprises work to integrate AI-enabled solutions into their networks, enterprises must not lose sight of the implications of these integrations and the added value they are working to unlock. It starts with understanding these capabilities and how they enable businesses to take their security infrastructure to the next level.
Doing more on the edge
The last two decades have seen technology vendors make significant strides in the computational capabilities of their hardware. Advances in chip architecture have led to multi-core processors and specialised graphics processing units (GPUs), meaning that processors can handle large amounts of data and execute more sophisticated algorithms.
These innovations trickle down and have an impact on all industries, including network security and surveillance. Edge computing enables video camera solutions to process and store data as close to the source as possible, rather than having to send it to a centralised hub. This also means that operators do not need to invest significantly in centralised infrastructure. Instead, they can optimise existing infrastructure with solutions that gather, process and store video data all at the same location; all while reducing network latency and bandwidth requirements.
In the end, as hardware costs continue to decrease, edge computing and AI-enabled devices for security applications can prove to be more cost-effective to businesses.
A is for analytics
Innovations in network cameras and video technology have enabled businesses to find what they are looking for much more easily. Previously, if there were a building intrusion or some sort of incident caught on camera, operators would have to sift through large amounts of raw footage to find the suspect, verify them, and then track their activities across any number of feeds.
AI-enabled analytics, which utilise algorithms to analyse live and recorded video content to generate descriptions of what is happening in a scene, enable camera operators to immediately identify objects, humans and vehicles. It also accelerates forensic investigations by letting operators pinpoint specific video points, as well as giving them the insights to improve building and overall business efficiency. For instance, operators can identify areas that experience high levels of foot and vehicle traffic, and subsequent congestion. They can then take steps to transform the space and its layout, thus ensuring that people and vehicles can move freely and in an optimised manner.
Being secure when transforming security
There is not one corner of the security industry that the AI trend is not impacting. As more security systems use AI-powered analytics and algorithms, organisations need to evaluate new and existing risks. They must understand the complex interplay between AI components and existing infrastructure, and catalogue all potential weak points, including those that threat actors and unauthorised users can exploit.
The concerns surrounding how threat actors leverage AI to carry out attacks also apply to video and security systems and their operators. In the same way that a company’s employees may be targeted by phishing attempts and ransomware, edge devices and network access points can also be targeted.
Video analytics will continue to benefit from the AI revolution, but enterprises must take a wholesale approach to the trend and integrate new technologies securely and diligently. In doing so, they can unlock the full benefits, while making their systems and infrastructure as resilient as possible.
© Technews Publishing (Pty) Ltd. | All Rights Reserved.
printer friendly version