Sentinel-2 Cloud Mask Annotations With Variability by Kaggle | Other

About this Dataset

Sentinel-2 Cloud Mask Annotations With Variability

Sentinel-2 Cloud Mask Annotations with Variability Tags

A Large, Representative Set of 20m Subscenes

By [source]

About this dataset

More Datasets

For more datasets, click here.

Featured Notebooks

🚨 Your notebook can be here! 🚨!

How to use the dataset

In this guide, we will cover how to use this dataset and what information can be derived from it.

First, let’s take a look at the columns in the dataset. We have scene name, difficulty level, annotator name, shadows_marked (yes/no), clear percent, cloud percent, shadow percent, dataset type (WorldView 2 or 3), forest/jungle coverage percentage details etc., snow/ice coverage percentage details etc., agricultural coverage percentage details etc., urban/developed coverage percentage details etc., coastal coverage percentage details etc., hills/mountains coverage percentage details etc., desert/barren coverage percentage details etc., shrublands/plains coverage percentage details(etc.), wetland/bog marsh coverage%, open water%, enclosed water%, thin cloud % , thick clouds % , low clouds % , high clouds % , isolated clouds % along with extended cloud type (altocumulus / stratocumulus) cirrus haze / fog , ice_clouds & contrails . All of these columns provide detailed percentages about different types of landcover along with corresponding cloud types & other useful information like annotator name involved in creating annotation for a particular scene .

The data within each column can then be used to derive further insights about any given Sentinel-2 subscene including landcover as well as various associated meteorological events such as precipitation and wind patterns which could enable specific decision-making applications like crop monitoring or urban development tracking in addition to understanding environmental impacts over large areas easily visible through satellite imagery. Furthermore, by analyzing this data combined with other standard atmospheric parameters such as wind speed & direction it is possible to track storm path direction by looking at cyclonic activity predicted by different conditions pertaining to satellite images gathered previously allowing accurate forecasting opportunity .

Research Ideas

Using the geographical attributes associated with each scene, this dataset can be used to categorize cultures based on their characteristics and geography.

This dataset can be used to better understand climate data, by looking at how cloud formations are distributed in a region and in relation to weather patterns.

This dataset can also help with machine learning projects related to object detection, as the cloud patterns and layout of the scenes can be seen as objects that algorithms should try to recognize or identify correctly while training

Acknowledgements

If you use this dataset in your research, please credit the original authors.
Data Source

License

License: CC0 1.0 Universal (CC0 1.0) - Public Domain Dedication
No Copyright - You can copy, modify, distribute and perform the work, even for commercial purposes, all without asking permission. See Other Information.

Columns

File: classification_tags.csv

Column name	Description
scene	Unique identifier for each subscene. (String)
difficulty	Difficulty rating of the subscene. (Integer)
annotator	Name of the annotator who classified the subscene. (String)
shadows_marked	Whether shadows were marked in the subscene. (Boolean)
clear_percent	Percentage of clear sky in the subscene. (Float)
cloud_percent	Percentage of clouds in the subscene. (Float)
shadow_percent	Percentage of shadows in the subscene. (Float)
dataset	Dataset the subscene was taken from. (String)
forest/jungle	Percentage of forest/jungle in the subscene. (Float)
snow/ice	Percentage of snow/ice in the subscene. (Float)
agricultural	Percentage of agricultural land in the subscene. (Float)
urban/developed	Percentage of urban/developed land in the subscene. (Float)
coastal	Percentage of coastal land in the subscene. (Float)
hills/mountains	Percentage of hills/mountains in the subscene. (Float)
desert/barren	Percentage of desert/barren land in the subscene. (Float)
shrublands/plains	Percentage of shrublands/plains in the subscene. (Float)
wetland/bog/marsh	Percentage of wetland/bog/marsh in the subscene. (Float)
open_water	Percentage of open water in the subscene. (Float)
enclosed_water	Percentage of enclosed water in the subscene. (Float)
thin	Percentage of thin clouds in the subscene. (Float)
thick	Percentage of thick clouds in the subscene. (Float)
low	Percentage of low clouds in the subscene. (Float)
high	Percentage of high clouds in the subscene. (Float)
isolated	Percentage of isolated clouds in the subscene. (Float)
extended	Percentage of extended
cumulus	Percentage of cumulus clouds in the subscene. (Float)
cumulonimbus	Percentage of cumulonimbus clouds in the subscene. (Float)
altocumulus/stratocumulus	Percentage of altocumulus/stratocumulus clouds in the subscene. (Float)
cirrus	Percentage of cirrus clouds in the subscene. (Float
haze/fog	Percentage of haze/fog in the subscene. (Float)
ice_clouds	Percentage of ice clouds in the subscene. (Float)
contrails	Percentage of contrails in the subscene. (Float)

Acknowledgements

If you use this dataset in your research, please credit the original authors.
If you use this dataset in your research, please credit .

Tables

Classification Tags

@kaggle.thedevastator_sentinel_2_cloud_mask_annotations_with_variabili.classification_tags

47.55 KB
513 rows
33 columns


CREATE TABLE classification_tags (
  "index" BIGINT,
  "scene" VARCHAR,
  "difficulty" BIGINT,
  "annotator" VARCHAR,
  "shadows_marked" BIGINT,
  "clear_percent" DOUBLE,
  "cloud_percent" DOUBLE,
  "shadow_percent" DOUBLE,
  "dataset" VARCHAR,
  "forest_jungle" BIGINT,
  "snow_ice" BIGINT,
  "agricultural" BIGINT,
  "urban_developed" BIGINT,
  "coastal" BIGINT,
  "hills_mountains" BIGINT,
  "desert_barren" BIGINT,
  "shrublands_plains" BIGINT,
  "wetland_bog_marsh" BIGINT,
  "open_water" BIGINT,
  "enclosed_water" BIGINT,
  "thin" BIGINT,
  "thick" BIGINT,
  "low" BIGINT,
  "high" BIGINT,
  "isolated" BIGINT,
  "extended" BIGINT,
  "cumulus" BIGINT,
  "cumulonimbus" BIGINT,
  "altocumulus_stratocumulus" BIGINT,
  "cirrus" BIGINT,
  "haze_fog" BIGINT,
  "ice_clouds" BIGINT,
  "contrails" BIGINT
);