/marker-detector-using-yolo-v3

Primary LanguageJupyter Notebook

YOLO_v3_Marker_Detector

This project is used to detect markers based on live video or any photo given by the user.

First, marker photos were taken and Labelimg was used to bound the markers in boxes in turn generating their coordinates.

The coordinates were trained in Google Colab using YOLO v3.

Configuration file of the training as well as weights were downloaded and used locally.

Finally, marker was detected with codes from test_on_video.

Steps to run

Master branch includes all codes file.

Clone this Repo

Create Virtual Enviroment and activate

python -m venv YOLO_env

Install requirements.txt

pip install -r requirements.txt

To detect marker on live vioeo run

test_on_video.py

To detect marker on Image run and give image path

test_on_photo.py