SIH 2023 - Micro-Vibration Analysis Platform
Computer vision system to amplify and analyze micro-vibrations in video for equipment damage detection - 3rd Rank at Smart India Hackathon 2023 (National Level).
Project Snapshot
Role
Team Lead
Organization
Smart India Hackathon 2023 (Software Edition)
Timeline
2023
Status
Completed
Deployment
Hackathon prototype
Tech Stack
Impact & Results
- 3rd Rank - National Level, Smart India Hackathon 2023
- Non-contact equipment damage detection using video
A computer vision system that amplifies and analyzes micro-vibrations in video to detect equipment damage. This project secured the 3rd Rank at the national level of the Smart India Hackathon 2023.
The Problem
Monitoring equipment health in industrial settings typically requires expensive physical contact sensors. A non-contact, vision-based approach was needed to detect anomalous vibrations indicative of equipment stress or damage without physical intervention.
The Solution
An intelligent vision pipeline that applies motion amplification to standard video feeds, making imperceptible vibrations visible and analyzable to alert on potential damage.
Architecture & Implementation
Motion Amplification: Utilized Eulerian Video Magnification techniques to amplify subtle, imperceptible vibrations in video frames.
Analysis: Developed a computer vision pipeline to analyze the amplified frames and detect anomalous patterns.
Alerting: Real-time notification system when stress or damage signatures are detected.
Data Used: Video recordings of operating industrial equipment.
Tech Stack: Python, OpenCV, TensorFlow, Eulerian Video Magnification.
Challenges & Solutions
Signal Isolation: Isolating micro-vibrations from background camera shake and environmental noise.
Performance: Running Eulerian Video Magnification algorithms in near real-time on standard hardware.
Time Constraints: Architecting and delivering a complete pipeline within the strict 36-hour hackathon timeframe.
Key Takeaways
Advanced computer vision and signal processing techniques (Eulerian Video Magnification).
Fast prototyping and system integration.
Effective team leadership and delegation under extreme time constraints.