Tilapia Fish Weight Estimation Using Deep Learning and Computer Vision

Introduction

This study presents a 2D computer vision approach to estimate Tilapia weight in freshwater using a multi-camera system for video and image capturing.

Methodology

The method relies on a custom Tilapia image dataset including various fish growth stages, involving two main steps:

  • Fish Detection: Using Mask R-CNN model for automatic Tilapia detection
  • Weight Estimation: Calculating depth to convert pixel dimensions to centimeters, followed by weight prediction using regression models

Models Tested

Three regression models were tested for accuracy:

  1. Linear Regression
  2. Random Forest
  3. Support Vector Regression (SVR)

Advantages

  • ✓ Cost-effective solution
  • ✓ No need for expensive stereo cameras
  • ✓ Non-intrusive, stress-free for fish
  • ✓ High accuracy across growth stages

Grand Innovative Ground launched a smart fish farming management system for Tilapia in freshwater, combining AI and computer vision.

📞 +968 91444528 | 📧 info@grandoman.net
📍 Bousher - Al Maha Street, London Square Building, 1st Floor, Office 101