Garbage Classification System

Automating waste sorting using deep learning and genetic algorithms

Tech Stack: Python TensorFlow MobileNet/Xception Genetic Algorithm SVM
Categories: Computer Vision Deep Learning Feature Selection

Project Impact

This system helps communities properly sort waste into 6 categories with 97.4% accuracy, promoting better recycling and environmental sustainability.

How It Works

1 Feature Extraction

Uses MobileNet/Xception CNN models to extract visual features from waste images

2 Genetic Optimization

Applies genetic algorithm to select most relevant features for classification

3 Final Classification

Uses SVM classifier on optimized features to determine waste category

97.4%
Training Accuracy
88%
Test Accuracy
6
Waste Categories

Technical Highlights

View on GitHub Research Paper