With the amount and quality of collected video data increasing every day, Video Analytics (VA) algorithms are gaining popularity in use cases that include situational awareness, post-incident investigation, real-time analysis and monitoring, and extraction of insights and trends for further analysis.
VA is already being used in counting the number of pedestrians to assess pathway effectiveness, counting the number of fish feed to optimise the operations of aquaculture hatcheries, and many other use cases.
Having worked with multiple agencies on a myriad of CCTV and VA requirements, GovTech’s Data Science & Artificial Intelligence Division (DSAID) developed the Video Analytics System (VAS) to address their needs across the Whole of Government (WOG).
As a central platform, VAS aims to:
- Enable agencies to explore the use of video analytics, develop and deploy VA solutions
- Exchange video data and analytics expertise across WOG
- Expand the video sensor network and insights derived to aid in operational workflows
With VAS, agencies can kickstart, deploy and develop VA projects quickly and effortlessly. They can leverage on their own videos as well as videos from other agencies, and make use of existing VA algorithms from GovTech and across the Government on Commercial Cloud (GCC) ecosystem. For custom use cases, they can develop their own VA models and still maintain operational flexibility vis-à-vis private commercial vendors.
The AI-powered VAS is part of the Service Layer of the Singapore Government Tech Stack(SGTS).
Speeds up VA tasks: Agencies with VA use-cases can benefit from the get-go with the accessible and intuitive system. Users will appreciate the ability to quickly generate insights and analysis from videos collected during trials, proof of concepts, and pre-analysis work.
Reduces manual analysis: With VAS’ automated analysis processes, agencies can unlock critical time and cost savings. For example, to help fish hatcheries optimise the supply monitoring of rotifers (a type of microorganism used as fish feed), officers from the Singapore Food Agency (SFA) had previously relied on the manual counting of rotifers, which can number around 1,000 per sample. With the help of VAS, SFA officers now spend considerably less time on laborious manual counting.
Designed for agencies’ needs: By working closely with agencies on their projects and consulting them through various Design Thinking workshops, DSAID has designed VAS with a focus on the VA needs of agencies, from requirements in the Instruction Manual for ICT&SS Management (formerly known as IM8) to integration and operational considerations.
In response to the COVID-19 pandemic, Singapore implemented Circuit Breaker measures in 2020 to contain the spread of the virus. These measures included the closure of non-essential workplaces and the prohibition of dining-in at eateries. During this Circuit Breaker period, many Singaporeans sought respite in parks and green spaces. As a result, some popular parks were congested.
In response, NParks and GovTech collaborated to develop Safe Distance @ Parks, a website that provides real-time information on crowd density at Singapore’s green spaces. It allows users to check the crowd density before visiting a park. It also allows the authorities to monitor and prevent the parks from being too crowded.
VAS is a crucial component of Safe Distance @ Parks. It can be understood as the link between the physical hardware and the Safe Distance @ Parks website. Input from CCTV cameras physically installed in the parks are sent to VAS for analysis. Thereafter, the analysed information is fed to and displayed on the Safe Distance @ Parks portal.
Reach out to the product team with your queries or feedback through this form.
Last updated 20 February 2023
Thanks for letting us know that this page is useful for you!
If you've got a moment, please tell us what we did right so that we can do more of it.
Did this page help you? - No
Thanks for letting us know that this page still needs work to be done.
If you've got a moment, please tell us how we can make this page better.