This project aims to detect PPE safety kits (Personal Protective Equipment) using YOLO (You Only Look Once) object detection algorithm. The model can detect various types of equipment including 'Hardhat', 'Mask', 'NO-Hardhat', 'NO-Mask', 'NO-Safety Vest', 'Person', 'Safety Cone','Safety Vest', 'machinery', 'vehicle' .The target audience for this project includes individuals or organizations who want to ensure the safety of their employees or people working in hazardous environments.
We used the PPE detection dataset available on Roboflow for training our model. The dataset contains images of people wearing PPE equipment in different environments.Dataset is annotated and have classes = ['Hardhat', 'Mask', 'NO-Hardhat', 'NO-Mask', 'NO-Safety Vest', 'Person', 'Safety Cone','Safety Vest', 'machinery','vehicle'] . Training set contains 2.6k images , validation set contains 114 images and Testing set contains 82 images. Input dimensions of single image is 640x640.
- OpenCV
- Ultralytics
- torch
- tochvision
- Python 3.10
- flask
- clone the repository :
$ git clone https://github.com/Vinayakmane47/PPE_detection_YOLO.git - create a conda environment :
$ conda create -n myyolo python=3.10 -y - Intall requirements :
$ pip install -r requirements.txt - Run main.py :
python app.py - Upload Image/Video to detect object or Start Live videocam
