Our India-based client required a single dashboard to help safeguard homes, industries, and public areas, since it may be difficult to notice the presence of fire, smoke, or crowding when there are multiple cameras at work. Our client requested us to create a dashboard which would be able to detect these hazards and count people in that particular area.
The client wanted a working demonstration that spots fire and smoke, counts people in each area and shows everything on one screen with alerts, a photo and the exact place, so we built the whole solution from the first idea to a tested system.
Our AI engineer taught the computer to recognise fire, smoke and people from live cameras, while our developers built the alert system and a live floor-plan screen with charts, our data engineer prepared the reports, our tester checked every situation and our project lead kept communication clear.
The main challenges were false alarms from steam and sunlight, wrong counts in crowds, repeated alerts and live updates from many cameras, which we solved by checking events over time, ignoring risky spots, joining repeated alerts into one and sending light updates, and during an alert the screen also shows how many people remain in each area, so the client now sees everything in one place, responds faster and keeps clear records for audits and reviews.
Python, ReactJS, PostgreSQL, PyTorch, Docker
YOLO Vision AI Object Detection for People, Chairs & Animals
View DetailsYOLO Vision AI Fire & Smoke Detection Solution
View DetailsChat bot developed in AI/ML for Fast Track Sourcing
View DetailsNBA Game prediction model using ML & Blockchain
View Details