Skip to main content

Python library for solving computer vision tasks specifically for satellite imagery

Project description

earth-vision

Earth Vision is a python library for solving computer vision tasks specifically for satellite imagery.

Objective

To ease researcher to run ML pipelines for AI or Deep Learning Applications in solving Earth Observation (EO) tasks.

Installation

We recommend Anaconda as Python package management system and using Python 3.9.

pip:

pip install earth-vision
conda install gdal

From source:

python setup.py install
conda install gdal

GDAL is actually a C++ library with python bindings. That means it relies on underlying C++ code and the package must be built/compiled in a certain manner to be usable with Python. So, we prefer to install it from Anaconda.

Example

from torch.utils.data import DataLoader
from torchvision.transforms import ToTensor, Compose, Normalize
from earthvision.datasets import RESISC45
from earthvision.models.resisc45 import regnet_y_400mf

# Transformation
preprocess = Compose([ToTensor(), 
                      Normalize(mean=[0.3680, 0.3810, 0.3436], 
                                std=[0.1454, 0.1356, 0.1320])])

# Dataset and Dataloader
dataset = RESISC45(root='../dataset', transform=preprocess, download=True)
dataloader = DataLoader(dataset, batch_size=32, shuffle=True)

# Model
model = regnet_y_400mf(pretrained=True)

Features Plans

Feel free to suggest features you would like to see by opening an issue.

  • GPU memory optimization [TBD]
  • High-level pipeline to integrate varied data sources [TBD]

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

earth-vision-1.0.0.tar.gz (43.2 kB view details)

Uploaded Source

File details

Details for the file earth-vision-1.0.0.tar.gz.

File metadata

  • Download URL: earth-vision-1.0.0.tar.gz
  • Upload date:
  • Size: 43.2 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.3.0 pkginfo/1.7.0 requests/2.22.0 setuptools/45.2.0 requests-toolbelt/0.9.1 tqdm/4.56.0 CPython/3.8.10

File hashes

Hashes for earth-vision-1.0.0.tar.gz
Algorithm Hash digest
SHA256 153b55666cb2365e4edb7cf6cea4783c9d48399eb0a8f728c83fd09a796d5aaa
MD5 264cd41a28b54bd12e902320858c1cff
BLAKE2b-256 83bfeeea13c9ff33159cdc11bf3a2c062490d6460ae06bfe004c47ad19dc3f03

See more details on using hashes here.

Provenance

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page