5 steps small cities can take to become ‘Smart Cities’
Lizzi Goldmeier of BriefCam shares five key ways that cities can use their existing infrastructure to work towards a ‘smart’ future.
Every new year brings with it its own potential and if 2020 is anything to go by, 2021 will certainly have a flavour all of its own. In the cutting-edge world of video content analytics, the senior leaders at BriefCam have some thoughts and predictions for this exciting area of imaging technology.
The Growing Importance of Video Analytics
One thing is abundantly clear; video analytics software continues to be widely embraced by a variety of industries and plays an increasingly important role in ensuring business continuity and growth. For these reasons, we see many major Video Management Systems (VMS) manufacturers adding analytics capabilities to their offerings, acquiring analytics technologies and integrating with industry-leading providers. This trend will continue into 2021 and beyond.
Discover more: Watch a BriefCam platform overview
Deep Investment in Deep Learning and Advancing toward Predictive Analytics
Deep Learning (a subset of Artificial Intelligence) will continue to drive advances in many areas, such as medical imaging, robotics, computer graphics and big data. In the video analytics space, Deep Learning will be the main driver to improve accuracy, reduce costs, and expand video understanding. Computer Vision technology will continue to support Deep Learning technologies by efficiently solving simple tasks.
I expect to see more prediction-based capabilities, such as true anomaly detection derived directly from the video or from the rich video metadata generated. People are limited in the amount of data they can absorb and analyse, but AI can bridge those gaps, and metadata can be combined with other data sources and sensors for generating unprecedented business intelligence out of all that Big Data. As more data is aggregated over time and across multiple cameras, there will be significant advancement in dimensions that are more predictive.
These developments will make it easier for video analytics systems to detect more granular actions and interactions (falling, tripping, etc.) and automate alert configuration when norms and benchmarks are violated. Anomalous behaviour can be reviewed interactively on-demand for deep insights or in real-time to drive immediate alerting and response when time is of the essence and situations are evolving.
How Deep Learning Algorithms are Used in Video Analytics
Using Video Content Analytics to Track Trends, Identify Anomalies and Accelerate Real-Time Response
Video Analytics at the Edge and Innovation in the Face of COVID 19
From a technology perspective, the trend towards edge appliances and analytics is going to continue to grow tremendously and quickly. This transition to edge computing is part of the larger digital transformation and IoT trend, with IP cameras presenting an edge opportunity within the physical security industry. We see more cameras being manufactured with analysis capabilities, as well as the availability of AI chips that can drive computing at the edge at a lower cost than ever before. We might also see more edge appliances as a transition platform to leverage existing cameras in the field until a time when on-camera analytics are the industry standard. Coupled with these technology trends, we can expect further adoption of subscription-based license models, as their natural commercial extension.
Because the Covid 19 pandemic will continue to slow down the global economy well into 2021, it will become even more critical for organisations to leverage existing technology investments to get more value from them. For many industries, video analytics is a key enabler for business continuity, as these organisations must rely on technology to meet customer expectations and comply with government mandates for health and safety. Not only has Covid 19 validated the need for broad analytics as well as real-time video processing for immediate response, it has also affected the types of analytics that businesses need. In that sense, it's becoming clear that, although point solutions are helpful for solving individual pain points, a comprehensive platform approach is becoming increasingly important for video content analytics to support a variety of analytics, including people counting, proximity detection, and face mask detection.
Emerging Infrastructure: The future of video analytics
Video content analytics for combating the public health crisis