Factory Safety Pose and Helmet Recognition
A factory safety prototype combining OpenPose, YOLO, and CNNs for human, helmet, and hazardous-pose recognition.
OpenPoseYOLOCNNComputer Vision
Use case
This project explored real-time vision for detecting hazardous poses and safety risks in factory environments.
System and methods
OpenPose provides pose information, while YOLO and convolutional neural networks support human and helmet recognition.
My contribution
I trained human and helmet recognition models, built datasets, and tuned parameters to improve recognition in factory scenarios.
Outcome
Honorable Mention at the 6th Hand-in-Hand Intelligent Innovation Application Competition.