Overview

Precision Aquaculture: An Integrated Computer Vision and IoT Approach for Optimized Tilapia Feeding

This repository contains a YOLOv8-based model for precise Tilapia feeding in aquaculture, combining computer vision and IoT technologies. Our system uses real-time IoT sensors to monitor water quality and computer vision to analyze fish size and count, determining optimal feed amounts. We achieved 94% precision in keypoint detection on a dataset of 3,500 annotated Tilapia images, enabling accurate weight estimation from fish length. The system includes a mobile app for remote monitoring and control. Our approach significantly improves aquaculture efficiency, with preliminary estimates suggesting a potential increase in production of up to 58 times compared to traditional farming methods. This repository includes our trained models, code, and a curated open-source dataset of annotated Tilapia images.

[Rest of the README content remains the same]

How to use

Please download the model weights first

Counting Model

Keypoint Detection Model

Paper

from ultralytics import YOLO
from PIL import Image

img = Image.open('<image-path>')
model = YOLO('<weights-path>')
results = model(img)

Results

Applications

This fish counting model can be useful in various scenarios, including:

Monitoring fish populations in aquariums or fish farms
Ecological studies in natural water bodies
Automated fish stock assessment

Citation

If you use this model in your research, please cite:

@article{hossam2024precision,
  title={Precision Aquaculture: An Integrated Computer Vision and IoT Approach for Optimized Tilapia Feeding},
  author={Hossam, Rania and Heakl, Ahmed and Gomaa, Walid},
  journal={arXiv preprint arXiv:2409.08695},
  year={2024},
  doi={10.48550/arXiv.2409.08695}
}