Switch to otb for direct grid

CHANGELOG.md

@@ -1,6 +1,6 @@
 # Sen2VM Release history
 
-## Merged in main
+## Merged in main (1.3.0.rc2)
 
 ### Main feature
 
@@ -12,11 +12,17 @@
 
 * Feature:
 
-  * 
+  * Handle mosaic DEM (multiple square degrees): #62
 
 * Doc:
 
-  * 
+  * Notebooks improvements:
+
+    * Separate static input download in a dedicated notebook (#56)
+    * Allow more flexible input structure with separated input locations (#56)
+    * Creation of a Notebook for Inverse location (#54)
+    * Allow using OTB instead of gdal in direct location Notebook (through a new notebook)
+
 
 ## 1.2.0 (2026-05-04)
 

Dockerfile

@@ -1,6 +1,6 @@
 FROM ghcr.io/sen2vm/sen2vm-build-env:latest AS launcher
 
-ENV SEN2VM_VERSION=1.3.0.rc1
+ENV SEN2VM_VERSION=1.3.0.rc2
 
 WORKDIR /Sen2vm
 

pom.xml

@@ -7,7 +7,7 @@
     <groupId>esa.sen2vm</groupId>
     <artifactId>sen2vm-core</artifactId>
     <!-- NOTE: remind to change also the SEN2VM_VERSION in the Dokerfile-->
-    <version>1.3.0.rc1</version>
+    <version>1.3.0.rc2</version>
 
     <name>sen2vm-core</name>
     <url>https://github.com/sen2vm/sen2vm-core</url>

sen2vm-notebook/README_Notebooks.md

@@ -16,44 +16,97 @@
 # Sen2VM Notebook Processing Workflow
 
 This repository provides a complete workflow to run **Sen2VM** inside Docker and generate orthorectified and mosaicked outputs from Sentinel-2 L1B data.  
+
+Four  notebooks are provided:
+
+* **inputs-download-notebook.ipynb**  
+  Downloads all required input data (GEOID and GIPP) from the Git repository.
+
+* **notebook-direct-grid.ipynb**  
+  Runs Sen2VM with the **direct** mode, using the downloaded inputs, along with the IERS file and L1B product, to produce orthorectified and mosaicked images using gdal functions.
+
+* **notebook-direct-grid-otb.ipynb**  
+  Runs Sen2vm with the **direct** mode, using the downloaded inputs, along with the IERS file and L1B product, to product orthorectified and mosaicked images using otb functions.
+
+* **notebook-inverse-grid.ipynb**  
+  Runs Sen2VM with the **inverse** mode, using the downloaded inputs, along with the IERS file and L1B product, to produce orthorectified and mosaicked images.
+
 The project is designed and tested on **Linux**. It may not work reliably on **Windows**.
-Actually, this notebook can generate an inverse grid but can not use it to apply the orthorectification. 
 
 ## Prerequisites
 
 * **Docker** is avaialble on the running machine (installed and running)
 * **Python 3.x** 
 * **the full sen2vm-notebook folder shall be present:**
 
-  * notebook.ipynb
+  * inputs-download-notebook.ipynb
+  * notebook-direct-grid.ipynb
+  * notebook-direct-grid-otb.ipynb
+  * notebook-inverse-grid.ipynb
+
   * requirements.txt
   * gdal-latest folder
 
 * Required data:
 
-  * **Sentinel-2 L1B product** : /DATASTRIP and /GRANULE.  
-  * **DEM files** placed in the appropriate directory
+  * **Sentinel-2 L1B product**: /DATASTRIP and /GRANULE.  
+  * **DEM files**
 
 ## Mandatory Directory Structure
 
+Only a `WORKDIR` folder is required. All outputs, as well as intermediate files generated during execution, will be stored there.
+
+## Inputs
+
+Several inputs are needed:
+* L1B S2 product under EUP.SAFE format:
+```bash
+ <L1B_product>/
+    ├── DATASTRIP  # Required
+    ├── GRANULE   # Required
+    ├── S2*OPER_MTD_SAFL1B_PDMC*.xml #Required
+    └──  ...
+```
+* Digital Elevation Model (DEM)
+
+* IERS (prediction of Earth orientation, IERS can be donwloaded using **notebook.ipynb** (cell number 2))
+
+* GIPP (GIPP can be downloaded from the GIT repositery using **inputs-download-notebook.ipynb**)
+
+* GEOID (GEOID can also be downloaded from the GIT repositery using **inputs-download-notebook.ipynb** )
+
+Thanks to a recent update (#issue 56), now the differents inputs path are absolute path.
+
+## Directory Structure after execution
 ```bash
 WORKDIR
 │
-├── DATA/
-│ ├── DEM/ # Put your DEM files here
-| ├── GEOID/ # Put your GEOID files here (Optional)
-│ └── <L1B_product>/ # Place the full L1B product here
-|   ├── DATASTRIP  # Required
-|   ├── GRANULE   # Required
-|   └──  ...
-└── ...
+├── bulletin*.txt # IERS prediction of earth exploration (downloaded by the cell n°3)  
+│
+├── output/ # output after orthorectification and mosaic (.tif)
+│ ├── DIRECT_gdal
+│ │     ├─ GDAL_OUTPUT_ORTHO  # .tif orthorectified, one for each band and detector
+│ │     └─ GDAL_OUTPUT_MOSAIC # .tif orthorectified and merged by band, one for each band
+│ ├── DIRECT_otb
+│ │     ├─ INVERSE_GRID  # inverse gelocation grid, generated from direct grids by sen2vminvlocfromdirlocgrid.py, these grids encode the transformation between sensor geometry and map geometry
+│ │     ├─ raw           # contains the raw L1B images extracted per detector and per band, expressed in the sensor geometry (i.e., before any orthorectification or map projection)
+│ │     ├─ otb_no_georef # contains images resampled using OTB with the inverse grids. The data are projected onto a regular grid but are not yet fully georeferenced
+│ │     ├─ output_georef # contains the final orthorectificatied and georeferenced images
+│ │     └─ mosaic        # contains mosaicked images generated per spectral band by merging all detector-level orthorectified 
+│ └── INVERSE
+│       ├─ INVERSE_GRID  # inverse gelocation grid, generated by sen2vm, these grids encode the transformation between sensor geometry and map geometry
+│       ├─ raw           # contains the raw L1B images extracted per detector and per band, expressed in the sensor geometry (i.e., before any orthorectification or map projection)
+│       ├─ otb_no_georef # contains images resampled using OTB with the inverse grids. The data are projected onto a regular grid but are not yet fully georeferenced
+│       ├─ output_georef # contains the final orthorectificatied and georeferenced images
+│       └─ mosaic        # contains mosaicked images generated per spectral band by merging all detector-level orthorectified images
+│
+├── src/     # contains a .sh used to run in the gdal docker generated by the notebook
+└── UserConf # contains .txt config files generated by the notebook
 ```
 
-If the GEOID folder is empty, the notebopok will automaticaly use the geoid provided by sen2vm-core
-
 ## Python Environment Setup
 
-In the notebook directory create your venv :
+In the notebook directory create your venv:
 
 ```bash
 python3 -m venv .venv
@@ -63,38 +116,85 @@ pip install -r requirements.txt
 
 Select the virtual environment kernel in your Jupyter session.
 
-## Notebook Configuration
+## Notebooks Configuration
+
+In **inputs-download-notebook.ipynb**:
+In the first cell of the notebook:
+
+1. Set the paths to:
 
+   * The directory where you want the GIPP files to be downloaded  
+   * The directory where you want the GEOID files to be downloaded  
+
+In **notebook-direct-grid.ipynb** or **notebook-direct-grid-otb.ipynb** or **notebook-inverse-grid.ipynb**:
 In the first cell of the notebook:
 
-1. Set the path to:
+1. Set the absolute paths to:
 
    * The working directory 
    * The L1B product
-   * The output directory
+   * The GIPP directory
+   * The GEOID directory
+   * The DEM directory
+   * The IERS file (if you do not have the IERS file, put `""`, and one will be automatically  ed to the working directory)
+   * NB : The output directory will be created in the working directory 
 
 2. Adjust configuration parameters for:
 
    * Sen2VM
    * Orthorectification settings
+
+3. Specify whether the docker images should be removed 
 
 ## Processing Steps
 
-Execute the notebook cell by cell in the following order:
+If you do not have your own GIPP and GEOID files, or if they are not already downloaded,
+execute the **inputs-download-notebook.ipynb** cell by cell in the following order:
 
-1. Variable definitions  
+1. Path definitions
 2. Clone `sen2vm-gipp`, and manage GIPP assets  
-3. Automatic download of the IERS bulletin 
-4. Generation of `config.json` in: `/WORKDIR/UserConf`
-5. Generation of `params.json` in: `/WORKDIR/UserConf`
-6. Execution of sen2vm inside Docker  
-7. Generation of a `.sh` script, then execution inside a second Docker container running the latest GDAL:
-
-    * Orthorectification by band  
+3. Copy `Geoid` folder from `"/sen2vm-core/src/test/resources/DEM_GEOID"`
+
+Then if you want to use the direct grids with gdal functions only, execute the  **notebook-direct-grid.ipynb** cell by cell:
+
+1. Variable and path definitions 
+2. Automatic download of the IERS bulletin if none is provided by the user
+3. Generation of `config.json` in: `/WORKDIR/UserConf`
+4. Generation of `params.json` in: `/WORKDIR/UserConf`
+5. Execution of sen2vm inside Docker to generatre the direct grids 
+6. Generation of a `.sh` script, then execution inside a second Docker container running the latest GDAL:
+
+    * Orthorectification by band that generate images in `WORKDIR/DIRECT/GDAL_OUTPUT_ORTHO` using the function **gdalwarp**
+    * Mosaicking by band that generate an image in `WORKDIR/DIRECT/GDAL_OUTPUT_MOSAIC` using the function **gdalmerge.py**
+
+If you want to use the direct grids with otb functions for the orthorectification, execute the  **notebook-direct-grid-otb.ipynb** cell by cell:
+
+1. Variable and path definitions 
+2. Automatic download of the IERS bulletin if none is provided by the user
+3. Generation of `config.json` in: `/WORKDIR/UserConf`
+4. Generation of `params.json` in: `/WORKDIR/UserConf`
+5. Execution of sen2vm inside Docker to generate the direct grids
+6. Generation of inverse grids in `WORKDIR/output/INVERSE/INVERSE_GRID` from the direct direct grids using **sen2vminvlocfromdirlocgrid**
+7. Orthorectification using OTB inside a docker then mosaicking using GDAL:
+
+    * Orthorectification by band 
     * Mosaicking  
 
-Docker images are automatically cleaned up after each execution.
+Docker images can be removed or kept depending on the value of `REMOVE_DOCKER_IMAGE` in step 1. 
+
+If you want to use the inverse grids, execute the **notebook-inverse-grid.ipynb** cell by cell:
+
+1. Variable and path definitions 
+2. Automatic download of the IERS bulletin if none is provided by the user
+3. Generation of `config.json` in: `/WORKDIR/UserConf`
+4. Generation of `params.json` in: `/WORKDIR/UserConf`
+5. Execution of sen2vm inside Docker to generate the inverse grids in `WORKDIR/output/INVERSE/INVERSE_GRID` 
+6. Orthorectification using OTB inside a docker then mosaicking using GDAL:
+
+    * Orthorectification by band 
+    * Mosaicking  
 
+Docker images can be removed or kept depending on the value of `REMOVE_DOCKER_IMAGE` in step 1.
 ## Execution
 
 Run all notebook cells sequentially.