Pytorch_Tree-Species-Mapping

This project implements deep learning models for tree species mapping using PyTorch. It enables the automatic classification of tree species and the generation of spatial distribution maps.

Project Overview

The purpose of this project is to utilize deep learning models to classify tree species from sentinel-2 time series and create distribution maps. Core functionalities include:

• Preprocessing and loading tree species image datasets
• Training deep learning models using PyTorch
• Evaluating and validating the tree species classification model
• Generating large-scale tree species distribution maps

Project Structure

├── .idea/                  # IDE configuration files
├── __pycache__/            # Python cache files
├── bases/                  # Core components
├── configs/                # Configuration files
├── models/                 # Model architectures
├── utils/                  # Utility functions
├── experiments/            # Experiment results
│   └── forestformer_0927/  # Example experiments
├── data_loader.py          # Data loading module
├── train.py                # Training script
├── evaluate.py             # Validation script
├── mapping.py              # Tree species mapping implementation
└── requirements.txt        # Environment dependencies

Prerequisites

Install Required Packages

Before running the code, make sure to install all necessary dependencies:

pip install -r requirements.txt

Usage Guide

Data Preparation

1. Prepare the training dataset, including tree species images and their respective labels.
2. Organize the dataset structure according to the requirements of the data loader.

Model Training

To train the model, run the following command:

python train.py

Model Validation

After training, evaluate the model's performance using:

python evaluate.py

Tree Species Mapping

Use the trained model to generate large-scale tree species distribution maps:

python mapping.py

Model Architecture

This project employs an optimized deep learning network ForestFormer tailored for tree species mapping tasks. The detailed architecture can be found in the models/ directory.

Experiment Results

Experiment results, including model weights, training logs, and evaluation metrics, are stored in the experiments/ directory.

License

This project is licensed under the MIT License.