AquaNET Application

A comprehensive fish health monitoring application with AI-powered disease classification and weight prediction features.

AquaNET Dashboard
Image Upload Interface
Disease Classification Results

Application Overview

AquaNET is a comprehensive web application designed for aquaculture professionals and fish farmers to monitor fish health through AI-powered analysis. The application combines secure user management with advanced machine learning capabilities to provide actionable insights about fish diseases and growth metrics.

Key Features

User Authentication

Secure login system implemented with Flask-Login and password hashing to protect user data and ensure privacy.

Secure Image Upload

Upload fish images in PNG, JPG, or JPEG formats with secure storage and validation to prevent malicious files.

Fish Disease Classification

Utilizes a pre-trained ResNet18 model (best_model.pth) to accurately classify various fish diseases from uploaded images.

Fish Weight Prediction

Predicts fish weight based on species and physical measurements (length, height, width) using machine learning algorithms.

MySQL Database Integration

Stores user data, fish information, and upload history in a structured MySQL database for efficient data management.

Upload History

Maintains a complete history of user uploads with results for easy reference and tracking of fish health over time.

Automatic Setup

Creates necessary database tables and an admin account automatically on first run for seamless deployment.

API Endpoints

Provides RESTful APIs for weight prediction and species retrieval to enable integration with other systems.

Project Information

  • Category: Web Application & AI
  • Technologies:
    Flask Python PyTorch MySQL ResNet18 Bootstrap JavaScript
  • Project Date: January 2025
  • Project URL: Live Demo
  • GitHub: Source Code

Technical Details

AquaNET is built with Flask as the backend framework, utilizing PyTorch for the machine learning components. The ResNet18 model was fine-tuned on a specialized dataset of fish diseases to achieve high classification accuracy. The application features a responsive frontend built with Bootstrap for optimal viewing on both desktop and mobile devices.

Challenges & Solutions

One of the main challenges was integrating the PyTorch model with the Flask application while maintaining performance. This was solved by implementing efficient image preprocessing and model loading techniques. Another challenge was ensuring the security of uploaded files, which was addressed through rigorous validation and secure storage practices.