Library for AI-Assisted Mapping Tool developed for Humanitarian OpenStreetMap Team

A small team from Omdena worked on a disaster management project. This package was created in order to simplify the integration of the data processing steps with the model training one.

`data` Directory Structure

.
├───images
│   ├───1
│   ├───2
│   ├───3
│   ├───4
│   └───5
├───inputs
│   ├───1
│   ├───2
│   ├───3
│   ├───4
│   └───5
├───masks
│   ├───1
│   ├───2
│   ├───3
│   ├───4
│   └───5
└───predictions
    ├───1
    ├───2
    ├───3
    ├───4
    └───5

inputs: GeoJSON labels and image files given to us.
images: Georeferenced images with the fourth band removed (if any).
masks: Rasterized labels.
predictions: Masks predicted by some ML model.

API Reference

preprocess(data_path, input_dir, image_dir, mask_dir)

Function for fully preprocessing the input data.
- data_path: Path of the directory where all the data are stored.
- input_dir: Name of the directory where the input data are stored.
- image_dir: Name of the directory where the images are stored.
- mask_dir: Name of the directory where the masks are stored.
predict(checkpoint_path, data_path, image_dir, pred_dir)

Function for predicting masks for all the input images.
- checkpoint_path: Path where the architecture and weights of the model can be found.
- data_path: Path of the directory where all the data are stored.
- image_dir: Name of the directory where the images are stored.
- pred_dir: Name of the directory where the predicted images will go.

Example Usages

Preprocessing:

from hotlib import preprocess

preprocess("/content/gdrive/MyDrive/Omdena/data", "inputs", "images", "masks")

Prediction:

from hotlib import predict

predict(
    "/content/gdrive/MyDrive/Omdena/checkpoints",
    "/content/gdrive/MyDrive/Omdena/data",
    "images",
    "predictions",
)

Release files for hotlib 1.0.33

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for hotlib 1.0.33
File	Size	Uploaded
hotlib-1.0.33.tar.gz	15.7 kB	Nov 29, 2022	Details

Built distribution (wheel)

Table of built distributions (wheels) for hotlib 1.0.33
File	Interpreter	ABI	Platform	Reset
hotlib-1.0.33-py3-none-any.whl 20.1 kB Nov 29, 2022	Python 3	none	any	Details

Total release size: 35.8 kB

Release files / hotlib-1.0.33.tar.gz

Download URL	hotlib-1.0.33.tar.gz
Size	15.7 kB
Tags	Source
SHA-256 checksum How to use checksums	`176c4388ffb06cacd98c50ca2e3d2135f1ff4a63254a27312a61cd8fd3a56247`
BLAKE2b-256 checksum How to use checksums	`aebb256b08943ad7dc3f0d332e6200e775d9fdeb91bb2e093a47ac0728e9109a`
Upload date	Nov 29, 2022
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`twine/4.0.1 CPython/3.9.13`

Release files / hotlib-1.0.33-py3-none-any.whl

Download URL	hotlib-1.0.33-py3-none-any.whl
Size	20.1 kB
Tags	Python 3
SHA-256 checksum How to use checksums	`73312fe5c1e6561e409227f029d1fddf9254778b33e08e53dba3341455d3960e`
BLAKE2b-256 checksum How to use checksums	`621de81036f9fae622bf22798f71737740e607d718db245f9accd547346fbba5`
Upload date	Nov 29, 2022
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`twine/4.0.1 CPython/3.9.13`