By Henrik Söderlund
The physical security industry agrees that video analytics has improved in recent years. Florian Matusek, Product Group Director, Genetec, believes people counting is a good example. The accuracy has increased and there are more possible applications today.
He says: “New technology such as deep learning has allowed manufacturers to increase accuracy dramatically. The scenarios where you could leverage it in the past were limited to top-down views with few people walking through. Today, we can also handle high-volume people counting and can even leverage tilted views from already installed surveillance cameras”.
Better cyber security
Roy Band, Community Manager, Middle Europe, Milestone Systems, agrees the accuracy of video analytics has improved:
“But also the amount of data we can extract from live and recorded images. This data is the most valuable for users, because it allows them to search more efficiently, to create dashboards about specific data, to spot future trends based on data, and to act for prevention instead of merely reacting to an incident”, he says.
Koen Rommel, Senior Vice President Sales EMEA, Bosch Building Technologies, also thinks video analytic capabilities have developed and have had substantial impact on robustness (capable of functioning under extreme circumstances), customisation (introducing machine learning capabilities enabling devices to learn customer specific situations that need monitoring) and accuracy.
Thomas Lausten, CEO of Mobotix, claims that many video surveillance systems and applications are more cyber secure today than previously.
He also says: “Software analytics and artificial intelligence solutions are much more reliable and provide better analysis than first generation systems”.
A differentiated market
Jimmy Ek, Nordic Sales Manager at Axis Communications, says that the market for intelligent security systems is very differentiated today with various types of customer needs.
“Some end customers want simple solutions from one supplier, while others want customised solutions and a close dialogue with the manufacturer”, he says.
Thomas Lausten says that Mobotix has experienced a significant increased focus on cyber security compared to before when mainly features and price were requested.
He says. “System quality and reliability also seems to be more requested especially by many larger end users and this has increased demand in some vertical markets combined with increased usage of analytic integrations”.
The usage of edge-based systems has been predicted to increase and Thomas Lausten confirms this trend, including analytics at the edge securing a higher level of cyber security.
Customers want more data
Florian Matusek says the days are gone where customers buy a VMS, a business intelligence application and video analytics separately.
“Customers demand easier setup and use of these systems. Buying video analytics as a separate application is still being done but increasingly only in situations where specialised analytics are demanded that are not already available in a platform”, he says.
Florian Matusek also stresses that more and more customers want to leverage their existing camera network and get more data out of it, and not just for security reasons, but for operational use.
Koen Rommel also highlights the rising importance of data that is collected via multiple sensors and consolidated into actionable insights that can be used to improve overall security or support new opportunities like receiving live metrics about crowd density, store traffic etc.
Video analytics today and in the future
Florian Matusek sees a continued growth in demand for perimeter protection analytics and he believes more data from video analytics will be connected to other data in the future, since the true value of the data is created when this happens.
He says: “For example, can I combine access control and video to detect tailgating? Can I combine radar and video analytics to get more accurate perimeter protection?”
Roy Band thinks that the popularity of edge processing will grow, and more analytics will be processed at the edge and/or in the cloud.
Thomas Lausten sees customer demand from areas such as crowd management, face recognition (with and without masks) smoke and analysing behaviour based on age, gender etc.
“Video analytics will increase in quality and development will be even faster. Video management will be even more geared towards data analysis, and cyber security will become even more important to secure reliability and compliance with local laws and regulations combined with GDPR”, he says.
Jimmy Ek believes that the video analytics most in demand are those for perimeter protection and customer experience analysis. He stresses that in these times when ecommerce is growing, competition for customers is increasing and, stores and facilities will need more data about customers to be able to meet their expectations. Here, intelligent video technology and analytics can be helpful for both retailers and customers.