Plant-Classification-And-Disease-Recognition

Overview 🌱

This repository demonstrates the application of cutting-edge deep learning techniques for plant leaf classification, disease recognition, and segmentation. The goal is to empower agricultural and botanical research with AI-driven solutions for plant health monitoring and analysis.

The project utilizes pre-trained models fine-tuned for high performance, achieving remarkable accuracy in various tasks. It also explores innovative approaches like Siamese networks for one-shot learning, making it suitable for scenarios with limited labeled data.

Key Features

• Plant Leaf Classification:

Fine-tuned ViT, VGG, and MobileNet models to classify plant leaves, achieving a 99% test accuracy.

• One-Shot Disease Recognition:

Used Siamese networks to identify plant diseases in unseen images, providing a solution for cases with limited data.

• Plant Segmentation:

Fine-tuned SAM (Segment Anything Model) and U-Net for detailed and accurate plant leaf segmentation.