Dataset Setup

This tutorial walks through the process of labeling the TIFF dataset. The tutorial assumes the TIFF dataset is already extracted from Google Earth Engine.

⚠️ If you are running the labeling script: Activate the Conda virtual enviromnent!

---

Dataset Location in Project Repo

How to set up dataset for labeling

TREE

.
├── example_dataset
│   ├── full_img
│   └── label_img
├── README.md
├── requirements.txt
├── src
│   ├── checkpoint_data
│   ├── __pycache__
│   └── run_annotation.py
├── testing_data
│   ├── checkpoints
│   ├── data_loader.ipynb
│   ├── full_img
│   └── label_img
└── tests
    ├── polygon_mask_test.pickle
    ├── test_config.yml
    └── test_run_annotation.py

testing_data is the offical format for a (labeled) dataset

Example_dataset will be our example for this tutorial

Dataset Configuration

./example_dataset
├── full_img
│   ├── Engilchek_glacier_1999-07-09.tif
│   ├── Engilchek_glacier_1999-08-10.tif
│   ├── Engilchek_glacier_1999-08-26.tif
│   └── Engilchek_glacier_1999-10-13.tif
└── label_img
    ├── Engilchek_glacier_1999-07-09.tif
    ├── Engilchek_glacier_1999-08-10.tif
    ├── Engilchek_glacier_1999-08-26.tif
    └── Engilchek_glacier_1999-10-13.tif

⚠️ The full_img file name must be the same as the ones label_img

tip

subdirectory	description
full_img	This subdirectory is where the TIFF image file with ALL the bands are present
label_img	This subdirectory is where the TIFF image files that will be used to label (only consist of a few bands) are present