Everyone is talking about AI, and the surveillance industry is abuzz with the potential. With more products coming to the market, customers are considering the move to AI-enabled cameras and weighing up the benefits.
Axis has made a commitment to AI in our products. Fundamental to many of the AI and deep learning capabilities in our portfolio is the system-on-chip (SoC), ARTPEC-8. Our portfolio of AI-enabled cameras will be increasing in number over the coming months and years. Here we examine why it’s worth considering AI-enabled cameras as your next investment, not only in terms of future proofing, but for the benefits today.
Making decisions for the near and long-term future of surveillance solutions
Customers of security solutions need to think long-term – a surveillance camera can have a useful life of five years or more. Over this time a new camera needs to be seen as a platform for future use cases and value. While AI-enabled cameras bring very real benefits today, future platform innovations will accelerate the capabilities of AI-enabled cameras across every industry.
The role of AI, machine learning and deep learning in video surveillance is growing, enabling us to increasingly ‘teach’ our cameras to be far more intuitive about what is happening in the scene and analyzing it in real-time. If customers only focus on the AI-enabled analytics available today, they may miss the potential of the as yet undeveloped applications coming both in the near and long-term future.
While it might seem to some that an investment in AI-enabled cameras today is a leap into the unknown, there are many immediate benefits.
Benefits today, and the future’s even brighter
An investment in an AI-enabled camera will bring benefits to security, safety and operational efficiency from as soon as the camera is deployed. We already have a number of cameras in our portfolio that benefit from deep learning-based analytics, and more accurate object detection and classification is primary among them.
Currently, edge analysis of surveillance camera footage using video motion detection shows that something or someone is moving. Following this, and after further analysis by a video management system (VMS), a human operator interprets exactly what caused the motion and if it presents any threat or security risk. However, through machine learning and deep learning, we can ‘train’ a camera system to classify certain objects automatically, so that operators can focus on actual threats instead of swaying vegetation, fast-moving shadows, etc. The classified objects are sent as metadata to the VMS for further analysis by a human operator or used to trigger automated responses based on pre-set rules. As the initial analysis takes place on the camera – as opposed to a server or in the cloud – valuable bandwidth can be saved as only relevant data is sent through the system.
Metadata created alongside the video images – essentially data about the images – delivers a major leap forward. The potential is incredibly valuable for searching through vast amounts of video including vehicles and people - and the attributes associated with those objects, such as colors of vehicles and clothing or the direction of travel. Looking forward, it will also be central to spotting patterns and trends which will be invaluable in operational efficiency and planning.
The greater accuracy of edge analytics – and the ability to distinguish between multiple classes of objects – immediately reduces the rate of false positives and false negatives. With that comes a related reduction in time and resources to investigate the false positives. More proactively, edge analytics can create a more appropriate and timely response. Distributing processing throughout the system lowers costs and enables a better user experience with greater customer value.
The constant enhancements in the Axis camera platform enable further development and improved native edge analytics as well as driving the creation of a new generation of analytics applications.
The potential of AI-enabled cameras is only limited by the imagination of developers. They are constantly looking to expand analytics capabilities. AXIS Camera Application Platform (ACAP) version 4 is designed to lower the threshold for developers moving to the edge, leading to a new generation of computer vision applications for the Axis AI-enabled camera platform.
A smart decision in smart technology
The move to AI-enabled cameras shouldn’t feel like a leap of faith – at most it’s a small step of imagination. The benefits not only become apparent in time, but customers can also reap the rewards now. Anyone that has experienced the benefit of fewer false alarms, made possible by a deep learning application, will never look back.
Surveillance cameras have a long lifecycle; an investment in AI-enabled cameras today will be a decision that unlocks value for years to come. Don’t hinder the capabilities of your video surveillance system tomorrow through a short-sighted decision today.