Construction Site Hazards Identification Based on Computer Vision and Deep Learning

This project is a partial fulfillment of the requirements for the degree of Master of Science in Computer Science.

Special Thanks for my supervisor Dr. Mona Abdelbaset Sadek Ali for her continuous support.

Model YOLO-v5 was taken from the original author Github repository

Getting Started

In the terminal, cd to the project file and install all the required packagaes by running the following line:

pip install -qr requirements.txt

you can run the detection model by running the following line:

python3 detect.py --weight PPE_HE.pt --source 0

--weight PPE_HE.pt is used to detect workers/PPE. you can change the weight to:

--source 0 runs detection from the webcam. you can change the value to images or videos.

Name		Name	Last commit message	Last commit date
Latest commit History 7 Commits
classify		classify
data		data
models		models
segment		segment
utils		utils
CONTRIBUTING.md		CONTRIBUTING.md
HE_weights.pt		HE_weights.pt
LICENSE		LICENSE
PPE_HE.pt		PPE_HE.pt
PPE_weights.pt		PPE_weights.pt
README.md		README.md
benchmarks.py		benchmarks.py
detect.py		detect.py
export.py		export.py
gui.py		gui.py
hubconf.py		hubconf.py
main.py		main.py
requirements.txt		requirements.txt
setup.cfg		setup.cfg
train.py		train.py
tutorial.ipynb		tutorial.ipynb
val.py		val.py
yolov5s.pt		yolov5s.pt