Needless to say, video analytics are an important element of video surveillance, helping make sense of what’s in the video. With another year starting, it’s important to discuss what some of the top trends in this particular market are.
Needless to say,
video analytics are an important element of video surveillance, helping make sense of what’s in the video. With another year starting, it’s important to discuss what some of the top trends are in this particular market.
Video analytics have been around for quite some time. Before, they were marred by various issues, failing to live up to people's expectations. That's why they were seen more as hype or a buzzword. Yet today, video analytics have become more accurate and advanced, prompting people to realize their benefits in security and non-security applications.
In security, for example, facial recognition is used to identify suspicious or blacklisted individuals; license plate recognition is used for the same purpose – to identify problematic vehicles. In non-security, applications range from retail to manufacturing to healthcare. In the latter category, for example, pose estimation can help identify a person falling in a senior care facility. In manufacturing, defects can be instantly identified by video analytics.
With advances in video analytics, all eyes are watching what’s in store for the technology in 2020. A recent
blogpost by BriefCam, authored by CTO Tim Edlund, identified some of the top trends for this year, and these are summarized as follows.
AI
As opposed to conventional rules-based algorithms,
AI or deep learning can help boost video analytics' accuracy, allowing them to identify objects effectively even when the image quality is poor. AI is also powerful when it comes to object classification. “Ongoing algorithm research and development continues to make the extraction and analysis of data increasingly efficient and accurate and the innovation in this area will progress as we move into 2020 and beyond,” Edlund said.
Real-time video processing
According to Edlund, because real-time video processing continues to provide significant benefit to end users – based on advances in the sophistication of real-time alerting – a top priority for organizations already leveraging video analytics is expanding the camera coverage for real-time video processing. “By expanding the scope of real-time video processing, users can trigger alerts for broader environments and also increase data aggregation and visualization into dashboards in real time to drive business insights based on continuous processing with integrated video data sources,” he said.
Facial recognition
Facial recognition has been cited as a top trend for various security companies. Indeed, with applications encompassing both security and non-security fields, facial recognition demands will only grow despite certain controversies surrounding it. “As higher resolution video and more efficient processing technologies evolve, face recognition solutions are becoming more effective and accurate,” Edlund said. “While the progress in this field has been significant, face recognition remains controversial and will continue to do so until regulations around its use become more definitive.”
High-resolution video
More and more, video is moving towards
4K and even 8K resolutions, offering crystal sharp images that can make video analytics even more effective. Yet there is a caveat. “The steep hardware requirements to support heavier video processing remain a barrier. As long as the cost of hardware is high, mass market adoption of higher resolution cameras will be somewhat stilted, but it’s definitely an evolution we’ll continue to monitor in the coming years,” Edlund said.
Cloud-based video analytics
Cloud offers a cost-effective solution for customers who opt to pay for video surveillance and the analytics tools as a monthly service rather than a capital investment. Further, cloud offers stronger security now. “Cloud services incur a fraction of the capital expenditures, procurement procedures and installation and deployment expenses of on-premises implementations, and are well-suited to serving the needs of a multi-site deployment,” Edlund said. “Cloud platforms now also boast robust cybersecurity.”