iamtekson / Deep-learning-for-satellite-imagery

This repo contains the code for satellite image analysis using deep learning.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Deep learning for satellite imagery

This repo contain the code and information about the deep learning on various satellite imagery.

It contain following items,

  1. Building detection

For the building detection, I have tested on this dataset. The dataset is mainly for the building damage assessment but I tried it with the building detection.

Building segmentation

  1. LULC classification

It has two sub-folders, one is for kaggle dataset and another is for numpy array format pre-processed dataset.

The input and the mask for the kaggle dataset looks like below image,

image

The inuput and the mask for numpy array dataset looks like below image,

LULC segmentation

  1. Landslide mapping

Landslide pre, post image

About

This repo contains the code for satellite image analysis using deep learning.


Languages

Language:Jupyter Notebook 99.9%Language:Python 0.1%