
Machine learning (ML) lets software learn patterns from data. Computer vision (CV) is the part of ML that understands images and video. Together they can turn ordinary camera feeds into useful information.
What computer vision can do
In general, computer vision can be used for work such as:
- Number plate recognition (ANPR): read vehicle number plates from camera images, for gate access, parking or enforcement.
- Vehicle counting: count and classify vehicles at junctions to understand traffic and plan improvements.
- Parking space detection: spot empty spaces from overhead cameras and guide drivers to them.
- Site monitoring: detect people or vehicles in restricted areas and send alerts.
Why eDruk
Computer vision depends on reliable cameras, networks and software working together. eDruk brings experience in all three: CCTV fibre networks for the Royal Bhutan Police, GPS vehicle tracking for RSTA and bank vehicle fleets, the mRSTA mobile app, and more than two decades of software development. We earlier researched number-plate recognition and related traffic applications; that research is not a live system today. We can help you assess whether computer vision fits your problem, run a pilot, and connect the results to your existing systems.
Getting started
- Define the question you want answered, such as how many vehicles pass or which spaces are free
- Check your cameras, lighting and network
- Run a small pilot and measure accuracy
- Scale up and integrate with dashboards, apps or alerts
See also: AI and automation · IoT and the DrukBees Cube
Talk to us
Tell us what you need, whether it is a new system, a network, an IoT idea or a second opinion on a tender. We will reply with practical next steps.
- Email: edruk@edruk.com.bt
- Phone: +975 2 332277
- Fax: +975 2 335228
- Head office: RTC Junction, Karma House, Ngabiphu, Thimphu 11001, Bhutan
- Also in Thimphu: Ata Khangzang Building, Room 47
- P.O. Box: 1620, Thimphu
- Office hours: Monday–Friday 10 AM–6 PM, Saturday 10 AM–2 PM
