In this article, Oliver Vellacott, CEO of IndigoVision, warns against setting too high an expectation for what video analytics can deliver.

Analytics will detect 'suspicious movement' from people walking along a street.  Analytics will detect terrorists walking round a hillside a mile away.  Analytics will pick an offender out from a sea of faces.  These are just some of the misconceptions about analytics today.  Was there ever a technology that was so 'over-promised and under-delivered'?

No. The reality is that analytics are still very much in their infancy.  Expectation management is at the heart of the issue - being realistic with end users about what can be achieved.  The fundamental problem is that as humans we do these tasks without even thinking about it.  We read license plates and recognise faces totally subconsciously.  It may have taken us years of learning during childhood to acquire these skills, but the fact is that we take them for granted.

Computers, on the other hand, lack even the basics of visual intelligence.  They can perform some video analytic functions reliably, but often only by severely constraining the application.  Qualification is everything, and setting the end user's expectations is absolutely vital.

Analytics today

So, what can be done today?  License plate recognition has been around a long time and is well proven, however it is still not 100% accurate.  Face recognition is notoriously difficult to perform reliably, and is extremely easy to fool by using disguises, and for them to work with any degree of accuracy an excellent headshot of the subject is required.

There are, however, some bread-and-butter analytics functions which CAN be performed reasonably well today.  Motion detection is the simplest, most basic form of analytics.  Many manufacturers support it but very few systems achieve sufficiently low false alarm rates to be useable.  A system, which generates anything in excess of 20 false alarms in a night, becomes ineffective, because all alarms very quickly become ignored or the motion detection gets switched off just to prevent the operator getting 'drowned in alarms'!  Aesop unwittingly foresaw the reality of most security systems when he wrote The boy who cried Wolf.

Congestion detection has evolved from basic motion detection.  When the density of humans or cars reaches a certain level an alarm is triggered.  Counter flow looks for objects which move 'against the flow', and is valued in applications like airport security.  Virtual tripwire is also a refinement of motion detection in that it triggers an alarm when someone or something 'breaks' a line that has been drawn in the image.  This is useful in applications such as large areas with 'no go zones', such as in factories.  People are allowed to happily move in 'free areas' but the system alarms as soon as any of them move out of these - see sidebar 'Analytics A to Z' (at end of this document).

With all these analytics applications, camera positioning, lens selection and lighting are absolutely critical.  Just changing the camera position can easily improve the analytics performance by an order of magnitude.  For example, in counter flow detection the algorithm has a far easier job if the camera is positioned pointing down so that it sees the area in 'plan view'.  If it is looking from a perspective view - with human bodies occluding one another, the process of tracking people becomes much harder, and expectations can once again become unrealistic.

Crowded marketplace

There are literally hundreds of companies touting analytics software.  As a manufacturer of complete IP Video solutions, IndigoVision gets approached by on average one analytics provider every week, asking us to integrate their product into our IP Video management platform.  Why so many?  Because any small software company can develop a suite of video analytics by buying a framegrabber, a powerful PC and writing some software.

These are then sold as separate, stand-alone systems, which sit 'next to' the main CCTV system.  Video is split from the matrix and fed into the analytics system.  This has the limitation that it is not truly integrated into the operation of the main CCTV system and therefore delivers only limited benefits.  Analogue just doesn't really support integrated analytics. DVRs do to a greater extent, but they still remain largely 'islands', i.e. not integrated, just as with analogue CCTV.

Integrated analytics - the IP video solution

IP-based video management systems provide the ideal platform for powerful analytics to be completely integrated into the system, making them a core and integral part of its operation.  Leading IP Video solutions support analytics that can be performed in two fundamental modes: live (to detect events as they occur) and post processing (to test various scenarios on recorded footage).

The optimum place to locate live analytics is obviously at the camera, as it is the only truly scalable solution and also doesn't use up network bandwidth.  Central real time processing will eventually run out of steam, whereas every camera can have dedicated processing.  For example a camera with built-in analytics can monitor scene activity and transmit only on specified events (e.g. a person moving the wrong way through airport security).  This reduces unnecessary video traffic on the network, thus reducing the bandwidth requirements and simply cannot be done with traditional analogue.

The optimum place to locate post-processing analytics is obviously on a central server, so that recorded video can be searched many times, with different parameters.  One of the biggest time wasters for operators used to be fast-forwarding and rewinding through VCR tape.  It improved with DVRs, but most are still basically a digital fast forward or rewind.  Analytics offers the potential to further transform this essential task by searching large amounts of recorded video for possible events and having the operator validate them.  Computers do what they are good at - locating possible events - and humans do what they are good at - verifying those events.


Setting expectations is everything.  It's important not to believe all the ridiculous hype and nonsense about what analytics can deliver.  In 30 years time it will probably be possible, but today it's all about making sure that what is achievable is done extremely well.

The animation below depicts a series of scenarios in which video analytics might be used.

Back to top

Dr Oliver Vellacott
IndigoVision Group plc

Download PDF version Download PDF version

In case you missed it

Security & Safety Things becomes Azena, underscores advances in smart camera platform development
Security & Safety Things becomes Azena, underscores advances in smart camera platform development

Security & Safety Things is announcing that it has rebranded to Azena, a new brand name that underscores the company’s corporate growth and leading-edge smart camera platform and positions it for the next chapter in its ambitious plans for redefining video analytics. With a growing slate of global customer and partner collaborations and expanding geographic coverage, Azena will continue to increase the value of its platform for systems integrators and end customers. More than 100 AI-enabled video analytics apps Since its market introduction in 2018, Azena has grown to more than 120 employees spread across its headquarters in Munich, its technology Innovation Accelerator facility in Pittsburgh, and another development hub in Eindhoven, The Netherlands, all supporting the Azena open platform for smart cameras.Integrators can flexibly add or change apps on one or multiple cameras as needed for their customers The Azena platform is comprised of an open operating system for cameras and an Application Store with nearly 100 Artificial Intelligence (AI)-enabled video analytics apps. It enables smart cameras to simultaneously run multiple apps directly on the device. Integrators can flexibly add or change apps on one or multiple cameras as needed for their customers and use any of the 15 cameras from six different manufacturer partners in a variety of form factors.“Systems integrators play a crucial role in connecting the video analytic edge devices on our platform into the larger system landscape for a truly data-driven approach to security, operational intelligence and automation,” said Hartmut Schaper, chief executive officer, Azena. “Our new identity as Azena positions us for improved name recognition and market presence as we continue to add functionality and the potential for expansion into new markets for our systems integrator partners.” More than 40 use cases in 25+ verticals The Azena Application Store features apps that address more than 40 different use cases in at least 25 different vertical markets, ranging from traditional perimeter security, retail loss prevention and occupancy management to stadium security and even the unique needs of aquaculture. Some examples of use cases include: One U.S. professional hockey team, the Pittsburgh Penguins, is using the Azena platform to monitor crowding at its stadium entrances, license plate recognition for more efficient stadium parking and heat mapping for improved layouts of its fan merchandise retail outlets. An oil drilling company is deploying smart cameras running the Azena OS so operations staff can remotely monitor any pumping disruptions in the oil fields. A chemical plant is monitoring its locations for the presence of smoke to enhance  workplace safety measures Collaboration with Proseguy Systems integrator Prosegur, one of the world’s largest security companies, has announced its collaboration with Azena to use analytics on the edge as part of its Security Operations Center as a service offering. By deploying more sophisticated analytics to measure activity or automatically verify alarms, incoming alarm traffic from customer sites can be prefiltered, reducing the number of alarms needing to be handled by human operators in the SOC, enabling a more appropriate response.Integrators will find a host of other new features in the Azena platformIntegrators will find a host of other new features in the Azena platform designed to leverage device management capabilities and remote access for diagnosis and maintenance to cameras on a customer site, using Azena’s digital twin architecture. Other benefits include: Ability to run all the analytics apps from the Azena Application Store on the video stream of existing IP cameras by means of a small appliance from one of the camera manufacturer partners, bringing AI to already installed video systems Wide range of integration options to connect VMS systems, dashboard software, access systems, other apps or other cameras to support the creation of sophisticated end-to-end solutions Option for integrators to build and deploy custom solutions with apps available only to them and their customers via the Azena Application Store Ability to securely and remotely connect to a customer camera without a VPN A new integration assistant that quickly builds middleware for custom integrations between Azena components and third-party software and hardware Opportunity to negotiate directly with app developers on bulk pricing Standardised terms of use that can be adopted by all applications in the Application Store

How soon will access control cards become extinct and why?
How soon will access control cards become extinct and why?

Since the advent of the physical security industry, access control has been synonymous with physical cards, whether 125 kHz ‘prox’ cards or the newer smart card alternatives. However, other credentials have also come on the scene, including biometrics and even smart phones. Some of these choices have distinct cost and security advantages over physical cards. We asked this week’s Expert Panel Roundtable: How soon will the access control card become extinct and why? 

Addressing the Internet Of Things (IoT) and challenges in device design using a comprehensive approach
Addressing the Internet Of Things (IoT) and challenges in device design using a comprehensive approach

As the number of connected devices increases worldwide, the ways that they are being used, designed, and tested have also expanded. The rise of connected devices is demanding engineers to harness the power of the internet of things, which is expected to hit 28 billion by 2025. A comprehensive approach to device design is needed more than ever to address the challenges that this rapid growth will bring. Why engineers should be using IoT technology in product design The demand for devices designed to use the Internet of Things (IoT) technology is increasing as more industries are finding expanded ways to put them into use. Industries such as healthcare, automobiles, and agriculture are becoming more dependent on cloud capabilities and are therefore in need of new devices able to connect to it. Due to this rise in demand, an increasing amount of devices are delivering a multitude of benefits both to consumers and companies. However, this new wave of products has led to a growing list of challenges for engineers as they are forced to address IoT tech in regards to connectivity, regulations, longevity, and security. Ways to use IoT in the development process Engineers are facing these new challenges along with the normal pressure of deadlines and test considerations. By approaching all of these issues from a comprehensive point-of-view, the solutions become clearer and new device capabilities can be born. Let’s look at the challenges individually as well as possible solutions for them. Improving connectivity IoT enables data to be transferred between infrastructure, the cloud, and devices, making the process smooth  Because IoT is based around connection, it’s no surprise that the primary challenge for engineers to overcome is the improvement of connectivity between devices. IoT enables data to be transferred between infrastructure, the cloud, and devices, so making this process as smooth as possible is crucial. The main challenges involved with connectivity have to do with development and product testing while meeting industry standards and best practices. Additionally, many companies lack the necessary equipment and technology to develop new IoT devices, which makes it difficult to create scalable prototypes and test new products. Suggested solutions To address the issue of not having the expertise and necessary tools for testing, we suggest outsourcing the prototyping and evaluation process instead of attempting to tackle this in-house. By doing this, you’re able to free up resources that would otherwise be needed for expensive equipment and qualified staff. Helping comply with regulations When working with devices that are connected across the world, there is a complex web of regulations and conformance standards that can lead to challenges for engineers. The necessity of complying with these regulations while also pushing to meet deadlines can be burdensome and lead to an increase in production time and expenses. Failure to comply with global and regional laws, as well as system and carrier requirements, can lead to fines and costly setbacks. This type of failure can destroy a company’s reputation on top of causing financial losses, often leading to the loss of business. Suggested solutions By testing the IoT device design and components early, engineers can address any pre-compliance issues that may arise. During the early stages of development, we suggest using scalable and automated test systems readily available in the marketplace. Improved communication with other devices New challenges arise as new devices hit the market and existing technologies are redesigned to offer a better experience In the rapidly growing number of connected devices, new challenges will arise as new devices hit the market and existing technologies are redesigned to offer a better user experience. This rapid growth in devices will lead to congested networks leading to the necessity of devices being able to function in the midst of increased traffic and interference. Failure to do this will lead to delayed responses which could prove to be fatal. Suggested solutions The best solution for this issue is found in the evaluation process and supporting test methods that the Institute of Electrical and Electronics Engineers (IEEE) published in the American National Standard for Evaluation of Wireless Coexistence (ANSI). This process addresses the interconnectivity issues present in radio frequency environments. The outlined process involves defining the environment and evaluating the wireless performance of the equipment through thorough testing. An in-depth version can be found in its entirety online. Increasing the longevity of devices IoT devices are being used in vital industries such as healthcare and automotive so battery life and power consumption are two challenges that engineers must take seriously. A failure in this area could potentially lead to loss of life or safety concerns on the road. As new firmware and software are being designed to address these factors, engineers must be implementing them into IoT devices with the ability to be continually updated. Suggested solutions Longevity should be addressed in all aspects of the design process and tested thoroughly using a wide range of currents. By doing this, an engineer can simulate consumer applications to best predict performance. Security Security and privacy are concerns with any technology, but with the use of IoT in medical devices, it’s paramount Security has been a controversial issue for IoT since its inception. Security and privacy are concerns with any technology, but with the widespread use of IoT in medical devices, smart home appliances, and access control and surveillance, it’s paramount. For example, medical devices may store information about health parameters, medications, and prescriber information. In some cases, these devices may be controlled by an app, such as a smart pacemaker, to prevent heart arrhythmias. Naturally, a security issue in these devices could be devastating. Another example of dangerous security concern is with surveillance cameras and access control, such as for home or business security systems. These intelligent door locking systems contain locks, lock access controllers, and associated devices that communicate with each other. Suspicious activities are flagged with alerts and notifications, but if a hacker gains access, it can lead to real-world, physical danger. Security design points Here are some key points for security design: Physical security: IoT devices may be in external, isolated locations that are vulnerable to attack from not only hackers but by human contact. Embedding security protection on every IoT device is expensive, but it’s important for general security and data safety. Security of data exchange: Data protection is also important because data gets transmitted from IoT devices to the gateway, then onto the cloud. With surveillance and access control information or sensitive medical information, and encryption is vital to protecting data from a breach. Cloud storage security: Similar to data exchange, the information stored in medical devices, surveillance and access control systems, and some smart appliances with payment features, must be protected. This includes encryption and device authentication through access control, which can police what resources can be accessed and used. Update: Security vulnerabilities will always occur, so the key to addressing them is having a plan to address errors and release patches. Customers should also have options to secure devices quickly and effectively. Suggested solutions Engineers can include security and protection into IoT devices with early and perpetual testing throughout the design process. Most security breaches occur at endpoints or during updates, giving engineers a starting point for how to address them. Creating more secure devices Ensuring the security of connected devices should be of supreme importance for engineers as these devices are vulnerable to security breaches. The ultimate security of devices goes beyond the scope of engineering as the network and enterprise levels must also be secure to protect against potential threats. However, engineers play a role in this protection as well and should consider device security in the design process. Suggested solutions On a device level, engineers can help protect IoT devices from vulnerabilities by implementing early testing and continuing it throughout the design process. Most security transgressions occur at endpoints so this continual testing can, and should, create barriers to breaches. Regulations and compliance For IoT engineers, the complex web of regulations and compliance standards present new challenges Regulations and compliance surrounding data and technology are nothing new, but for IoT engineers, the complex web of regulations and compliance standards present new challenges. Engineers are already addressing obstacles in security and connectivity, all while meeting deadlines, and working around regulations adds time and expense to the process. Unfortunately, a failure to comply with global, regional, or local laws can lead to setbacks and fines. In addition to time lost in production and possible fines, the damage to a company’s reputation can lead to even more losses. Suggested solutions Compliance should be considered early and often in the design process. In the early stages of development, the IoT device or components can be tested to address and compliance issues. If possible, use a scalable and automated test system. The comprehensive solution As we stare at an uncertain future full of possibilities, it’s clear to see that new challenges will continue to be presented as technology evolves and new innovative devices are designed by engineers. By addressing these issues early and often, solutions can be implemented and problems prevented before they even have a chance to occur thanks to sound engineering and solid design.