Previous OGC Testbeds 13-16 initiated the design of an application package for Earth Observation Applications in distributed Cloud Platforms
+The application package provides information about the software item, metadata and dependencies
+Deployed and executed within an Exploitation Platform in a service compliant with the OGC API Processes specification
+OGC 20-089 Defines the Best Practice to package and deploy Earth Observation Applications in an Exploitation Platform
+OGC 20-089 defines guidance for the 3 viewpoints: +* For a Developer to adapt an application +* For an Integrator to package an application +* For an Platform to deploy and execute the application
+12 submitters organisations:
+Decouple application developers from exploitation platform operators and from application consumers:
+Enable exploitation platforms to virtually support any type of packaged EO application
+The Application Package:
+Application developers:
+The Common Workflow Language (CWL) is an open standard for describing analysis workflows and tools in a way that makes them portable and scalable across a variety of software and hardware environments, from workstations to cluster, cloud, and high performance computing environments.
+The CWL is a specification for describing analysis workflows and tools.
+CWL workflows are portable and scalable across a variety of software and hardware environments, from workstations to cluster, cloud, and high performance computing (HPC) environments.
+CWL is designed to meet the needs of data-intensive science, such as Bioinformatics, Medical Imaging, Astronomy, Physics, and Chemistry.
+ + + + + + +Purpose: to crop a particular band defined as a common band name (such as the "green" or "nir" band) from a satellite image acquired by either Sentinel-2 or Landsat-9.
+This step is highlighted below:
+The crop.py script is a command-line tool that takes as input
and then crops the specified band from the asset associated with the common band name to the specified AOI.
+It uses various Python libraries like pystac, rasterio, pyproj, shapely, and loguru.
Here is an overview of the script's functionality:
+It defines a function aoi2box to convert an AOI expressed as a bounding box string into a list of floats.
It defines a function get_asset to retrieve the asset of a STAC Item that is defined with a common band name. It iterates through the assets and checks if a band has the specified common name.
It defines a command-line interface using click, with options for providing the input STAC Item URL, AOI, EPSG code, and common band name.
The crop function is the main entry point. It reads the STAC Item specified by the input URL and retrieves the asset associated with the common band name. It then crops the asset to the specified AOI using the rasterio library.
It transforms the bounding box coordinates to match the EPSG code provided.
+It performs the cropping using the rasterio.mask.mask function.
It writes the cropped image to a GeoTIFF file with a filename like "crop_bandname.tif."
+The script is executable as a command-line tool as its usage is:
+Usage: app.py [OPTIONS]
+
+ Crops a STAC Item asset defined with its common band name
+
+Options:
+ --input-item TEXT STAC Item URL or staged STAC catalog [required]
+ --aoi TEXT Area of interest expressed as a bounding box [required]
+ --epsg TEXT EPSG code [required]
+ --band TEXT Common band name [required]
+ --help Show this message and exit.
+To use this script, you would typically run it from the command line, providing the necessary input options such as the STAC Item URL, AOI, EPSG code, and common band name:
+python app.py \
+ --input-item "https://earth-search.aws.element84.com/v0/collections/sentinel-s2-l2a-cogs/items/S2B_10TFK_20210713_0_L2A" \
+ --aoi "-121.399,39.834,-120.74,40.472" \
+ --epsg "EPSG:4326" \
+ --band "green"
+It will then crop the specified band from the STAC asset and save it as a GeoTIFF file.
+The Python code is provided here:
+| water-bodies/command-line-tools/crop/app.py | |
|---|---|
1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9 + 10 + 11 + 12 + 13 + 14 + 15 + 16 + 17 + 18 + 19 + 20 + 21 + 22 + 23 + 24 + 25 + 26 + 27 + 28 + 29 + 30 + 31 + 32 + 33 + 34 + 35 + 36 + 37 + 38 + 39 + 40 + 41 + 42 + 43 + 44 + 45 + 46 + 47 + 48 + 49 + 50 + 51 + 52 + 53 + 54 + 55 + 56 + 57 + 58 + 59 + 60 + 61 + 62 + 63 + 64 + 65 + 66 + 67 + 68 + 69 + 70 + 71 + 72 + 73 + 74 + 75 + 76 + 77 + 78 + 79 + 80 + 81 + 82 + 83 + 84 + 85 + 86 + 87 + 88 + 89 + 90 + 91 + 92 + 93 + 94 + 95 + 96 + 97 + 98 + 99 +100 +101 +102 +103 +104 | |
Purpose: to calculate the normalized difference of the "green" or "nir" bands.
+This step is highlighted below:
+The norm_diff script is a command-line tool for performing a normalized difference between two raster images.
It uses the click, rasterio, and numpy libraries to perform the calculation and save the result as a GeoTIFF file.
Here's an overview of what the script does:
+It defines a command-line interface using the click library, with two arguments for providing the file paths of the two raster images that you want to calculate the normalized difference for.
The normalized_difference function is the main entry point. It opens the two input raster files specified as arguments.
It reads the data from the first raster (specified by rasters[0]) using rasterio, and it also copies the metadata (e.g., projection, geotransform) of this raster to be used for the output.
It then opens the second raster (specified by rasters[1]) and reads its data.
It updates the data type in the metadata to "float32" because the normalized difference result will be a floating-point image.
It creates an output raster named "norm_diff.tif" using rasterio. This output raster will have the same metadata as the first input raster, but it will be of data type float32.
It calculates the normalized difference between the two input arrays (array1 - array2) / (array1 + array2) and writes it to the output raster using dst_dataset.write().
The script is executable as a command-line tool as its usage is:
+Usage: app.py [OPTIONS] RASTERS...
+
+ Performs a normalized difference
+
+Options:
+ --help Show this message and exit.
+The Python code is provided here:
+| water-bodies/command-line-tools/norm_diff/app.py | |
|---|---|
1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9 +10 +11 +12 +13 +14 +15 +16 +17 +18 +19 +20 +21 +22 +23 +24 +25 +26 +27 +28 +29 +30 +31 +32 +33 +34 +35 +36 | |
Purpose: to apply the Otsu threshold to the normalized difference.
+This step is highlighted below:
+The otsu Python script is a command-line tool for applying the Otsu threshold to a single input raster image.
It uses the click, rasterio, numpy, skimage.filters, and loguru libraries.
Here's an overview of what the script does:
+It defines a command-line interface using the click library, with a single argument for providing the file path of the input raster image on which you want to apply the Otsu threshold.
The otsu function is the main entry point. It opens the input raster file specified as the argument.
It reads the data from the input raster using rasterio and also copies the metadata (e.g., projection, geotransform) of this raster to be used for the output.
It applies the Otsu threshold to the input array by calling the threshold function. The threshold_otsu function from skimage.filters is used to calculate the Otsu threshold. The thresholding process marks pixels as True or False based on whether they are greater than the calculated threshold.
It creates an output raster named "otsu.tif" using rasterio. This output raster will have the same metadata as the input raster.
It writes the binary image to the output raster using dst_dataset.write().
The result, a binary image where pixel values are either True or False based on the thresholding, will be saved as "otsu.tif" in the same directory where the script is executed.
+The script is executable as a command-line tool as its usage is:
+Usage: app.py [OPTIONS] RASTER
+
+ Applies the Otsu threshold
+
+Options:
+ --help Show this message and exit.
+The Python code is provided here:
+| water-bodies/command-line-tools/otsu/app.py | |
|---|---|
1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9 +10 +11 +12 +13 +14 +15 +16 +17 +18 +19 +20 +21 +22 +23 +24 +25 +26 +27 +28 +29 +30 +31 +32 +33 +34 | |
Purpose: produce a STAC Catalog with a STAC Item describing the detected water body result.
+This step is highlighted below:
+The stac script is a command-line tool for creating a SpatioTemporal Asset Catalog (STAC) catalog containing detected water bodies.
It takes the STAC Item URLs and corresponding water body GeoTIFFs as input, creates STAC Items for each water body, and organizes them into a STAC Catalog.
+The script utilizes the click, pystac, rio_stac, and loguru libraries.
Here's an overview of what the script does:
+It defines a command-line interface using the click library. The script expects multiple input STAC Item URLs and corresponding water body GeoTIFFs as arguments.
The to_stac function is the main entry point. It creates a new STAC catalog and iterates through the provided STAC Item URLs and water body GeoTIFFs.
For each input pair (STAC Item URL and GeoTIFF), it reads the STAC Item, creates a directory with the same name as the item's ID, and copies the water body GeoTIFF into that directory.
+It uses the rio_stac library to create a STAC Item for the water body. This includes specifying the source GeoTIFF, input date/time, asset roles, asset href, and more.
The created STAC Items are added to the STAC catalog.
+After processing all input pairs, it saves the STAC catalog to the root directory, specifying it as a self-contained catalog with the catalog type set to pystac.CatalogType.SELF_CONTAINED.
The script will create a STAC catalog containing the detected water bodies and save it in the current directory.
+The script is executable as a command-line tool as its usage is:
+Usage: app.py [OPTIONS]
+
+ Creates a STAC catalog with the water bodies
+
+Options:
+ --input-item TEXT STAC Item URL [required]
+ --water-body TEXT Water body geotiff [required]
+ --help Show this message and exit.
+The Python code is provided here:
+| water-bodies/command-line-tools/stac/app.py | |
|---|---|
1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9 +10 +11 +12 +13 +14 +15 +16 +17 +18 +19 +20 +21 +22 +23 +24 +25 +26 +27 +28 +29 +30 +31 +32 +33 +34 +35 +36 +37 +38 +39 +40 +41 +42 +43 +44 +45 +46 +47 +48 +49 +50 +51 +52 +53 +54 +55 +56 +57 +58 +59 +60 +61 +62 +63 +64 | |
This application takes as input Copernicus Sentinel-2 or USSG Landsat-9 data and detects water bodies by applying the Otsu thresholding technique on the Normalized Difference Water Index (NDWI).
+The NDWI is calculated with:
+Typically, NDWI values of water bodies are larger than 0.2 and built-up features have positive values between 0 and 0.2.
+Vegetation has much smaller NDWI values, which results in distinguishing vegetation from water bodies easier.
+The NDWI values correspond to the following ranges:
+| Range | +Description | +
|---|---|
| 0,2 - 1 | +Water surface | +
| 0.0 - 0,2 | +Flooding, humidity | +
| -0,3 - 0.0 | +Moderate drought, non-aqueous surfaces | +
| -1 - -0.3 | +Drought, non-aqueous surfaces | +
To ease the determination of the water surface/non water surface, the Ostu thresholding technique is used.
+In the simplest form, the Otsu algorithm returns a single intensity threshold that separate pixels into two classes, foreground and background. This threshold is determined by minimizing intra-class intensity variance, or equivalently, by maximizing inter-class variance:
+
The application can be used in two modes:
+This scenario is depicted below:
+This scenario is depicted below:
+Alice packages the application as an Application Package to include a macro workflow that reads the list of Sentinel-2 STAC items references or Landsat-9 staged data, launches a sub-workflow to detect the water bodies and creates the STAC catalog:
+
The sub-workflow applies the crop, Normalized difference, Otsu threshold steps:
The development and test dataset is made of two Sentinel-2 acquisitions:
+| Acquisitions | ++ | + |
|---|---|---|
| Mission | +Sentinel-2 | +Sentinel-2 | +
| Date | +2022-05-24 | +2021-07-13 | +
| URL | +S2B_10TFK_20210713_0_L2A | +S2A_10TFK_20220524_0_L2A | +
| Quicklook | + |
+ |
+
And one Landsat-9 acquisition:
+| Acquisition | ++ |
|---|---|
| Date | +2023-10-15 | +
| URL | +LC09_L2SP_042033_20231015_02_T1 | +
| Quicklook | + |
+
Each Command Line Tool step such as crop, Normalized difference, Otsu threshold and Create STAC runs a simple Python script in a dedicated container.