+ . . PROJECT DETAIL . . +
Diesel Exhaust Fluid Dispensing Verification System
[ ACTIVE ]Computer vision pipeline for verifying DEF dispensing activity and preventing bag theft using CCTV feeds.
Developed a computer vision–based monitoring system to verify the physical dispensing of Diesel Exhaust Fluid (DEF) and prevent container theft using CCTV camera feeds. The system processes live RTSP video streams to detect DEF bags and dispensing activity, generating structured verification events for operational monitoring and auditing.
Key Contributions
- Implemented a YOLO-based object detection pipeline to identify DEF bags and dispensing activity from CCTV video streams.
- Integrated RTSP camera feeds to enable real-time monitoring of operational environments.
- Built an event-driven processing pipeline using Apache Kafka to stream detection and verification events.
- Stored detection events and verification logs in PostgreSQL to support auditing and operational analytics.
- Implemented logic to detect abnormal scenarios such as missing bags or removal events, enabling automated theft prevention alerts.
- Designed the system to handle multiple camera feeds and continuous real-time processing in production environments.
Tech Stack
Python, YOLO, OpenCV, RTSP, Apache Kafka, PostgreSQL, Computer Vision, Event Streaming
Links
Private – Developed for employer