The fact is, municipalities cannot practically apply traditional video recognition technology in highly variable and complex online situations, like a live city environment with its train stations, bus shelters, airports and other public areas. To anticipate what’s relevant based on the prescribed video surveillance task, the Nokia IoT Scene Analytics solution leverages:
Beyond enhancing situational awareness, the Nokia IoT Scene Analytics solution can also prioritize streams for their anticipated relevance, automatically allocating network resources based on dynamic application needs. The least relevant streams are cut off at the source ensuring network availability for the most critical ones. Existing video analytics solutions have some drawbacks, including false positives. These are caused because real scenes – like what cameras located in cities produce – are too complex for traditional object recognition software.
What is required is an intelligent platform that can proactively detect, select and track only relevant video streams for a variety of surveillance tasks. As a result, only a fraction of the massive set of video streams needed has to be handled, transmitted and stored. Using the Nokia IoT Scene Analytics solution, in parallel with the Nokia IMPACT IoT Platform, your video monitoring solution is agnostic to video resolutions, camera brands, and networking options.
Request More Information
Real-time monitoring and analytics
The Nokia IoT Scene Analytics solution, in effect, turns cameras into IoT sensors. The software derives abstract information vectors from each video feed (e.g., motion, direction, velocity, density, etc.), then uses machine learning technology – developed by Nokia’s Bell Labs – to analyze these vectors in real time, establishing patterns, identifying anomalies and generating alerts. The Nokia IoT Scene Analytics solution is better in real-life situations, where crowds of people and/or poor lighting limits visibility and where traditional rules-based solutions are inadequate.
For example, a camera that is monitoring a highway could be programmed only to transmit a video feed when unusual activities occur, such as slow-moving (or stopped) traffic, the presence of a pedestrian, an accident, or when a vehicle traveling in the wrong direction is detected. Another example could be a video feed from a public area, such as a street lined with shops.
"The Nokia IoT Scene Analytics solution leverages machine learning, pattern recognition, and behavioral knowledge to anticipate what’s relevant, based on the prescribed video surveillance task."