-
-
VoltVision AI GitHub repository structure and source code
-
Real-time electrical spark detection via OpenCV camera feed and VS Code environment
Inspiration
Electrical short circuits and spark hazards cause severe industrial fires and grid infrastructure damage. VoltVision AI was created to provide real-time, edge-based visual hazard detection to identify sparks and electrical anomalies before catastrophic failures occur.
What it does
VoltVision AI monitors live video streams from edge cameras. Utilizing OpenCV color thresholding and contour analysis in the HSV color space, it detects bright spark signatures and electrical fire risks. When a hazard is identified, the system visually highlights the area with a bounding box and triggers automated alert payloads formatted for AWS IoT Core integration.
How we built it
Computer Vision: Built with Python and OpenCV using HSV color space segmentation and contour detection algorithms. Cloud Architecture: Integrated Paho-MQTT client for secure communication and alert message publishing to AWS IoT Core. Version Control: Git and GitHub for modular source code repository management.
Challenges we ran into
Calibrating color detection thresholds to accurately isolate electrical spark signatures under varying ambient light conditions. Configuring TLS client certificate authentication for secure MQTT protocol transmission to AWS Cloud services.
Accomplishments that we're proud of
Successfully achieving real-time, low-latency spark detection on an edge video feed. Structuring clean, modular Python code ready for deployment and cloud synchronization.
What we learned
Effective application of HSV image filtering for real-time industrial safety monitoring. Setting up end-to-end edge-to-cloud IoT notification pipelines.
What's next for VoltVision AI
Training a custom deep learning model (such as YOLOv8) to expand detection beyond color-based thresholding. Full cloud deployment of AWS SNS email and SMS emergency alert workflows.
Log in or sign up for Devpost to join the conversation.