Weed Detector

Weeds in crops are every farmer's nightmare and removing them is a very time consuming and labor intensive job. This web app can assist farmer's in detecting weeds among acres of land with the motivation to save time and energy.

The idea was inspired from a Youtube video (see references) and Blue River Technology is a well known organisation for implementing such technology.

Demo

Live demo of the app hosted on streamlit:

Weed Detector Demo

Installation

Install using a Python virtual environment and pip. Need to clone the yolov5 repo as well

    git clone https://github.com/ultralytics/yolov5
    cd yolov5
    pip install -qr requirements.txt

Change the path of the weights in app.py

Deployment

To deploy this project run

    streamlit run app.py

Trained Weights

The trained weights can be found in releases

References

• AI & the future of Agriculture
• Crop and weed detection data with bouding boxes dataset