opencropphenotyping 0.2.0__tar.gz → 0.3.0__tar.gz

{opencropphenotyping-0.2.0/src/opencropphenotyping.egg-info → opencropphenotyping-0.3.0}/PKG-INFO

@@ -1,8 +1,11 @@
 Metadata-Version: 2.4
 Name: opencropphenotyping
-Version: 0.2.0
+Version: 0.3.0
 Summary: Open-source tools for crop phenotyping using multispectral geospatial imagery
 Author: Aurélien Peter
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Operating System :: OS Independent
 Requires-Python: >=3.11
 Description-Content-Type: text/markdown
 License-File: LICENSE
@@ -26,15 +29,23 @@ Dynamic: license-file
 
 [![CI](https://github.com/Aurelien-Peter/OpenCropPhenotyping/actions/workflows/ci.yml/badge.svg)](https://github.com/Aurelien-Peter/OpenCropPhenotyping/actions/workflows/ci.yml)
 [![codecov](https://codecov.io/github/Aurelien-Peter/OpenCropPhenotyping/graph/badge.svg?token=7SAC1M1VY5)](https://codecov.io/github/Aurelien-Peter/OpenCropPhenotyping)
+[![PyPI Version](https://img.shields.io/pypi/v/opencropphenotyping)](https://pypi.org/project/opencropphenotyping/)
+[![Python versions](https://img.shields.io/pypi/pyversions/opencropphenotyping)](https://pypi.org/project/opencropphenotyping/)
+[![Docker publish](https://github.com/Aurelien-Peter/OpenCropPhenotyping/actions/workflows/publish-docker.yml/badge.svg)](https://github.com/Aurelien-Peter/OpenCropPhenotyping/actions/workflows/publish_docker.yml)
 
 **An open-source Python framework for high-throughput crop phenotyping using drone and satellite imagery.**
 
 ## About the project
 
 OpenCropPhenotyping is an open-source project dedicated to reproducible high-throughput crop phenotyping using remote sensing data.
+
 The objective is to provide reproducible workflows to extract agronomic traits from UAV and Sentinel-2 imagery, from raw image preprocessing to trait extraction, statistical analysis and visualization. 
+
+
 The project currently combines multispectral Sentinel-2 processing with classical RGB UAV image analysis. The first UAV developments focus on crop-row detection, vegetation segmentation and plant-position analysis, with the objective of progressively extending the framework toward more advanced high-throughput phenotyping workflows.
 
+The project includes automated testing, continuous integration, code coverage monitoring, and package distribution through PyPI. Docker support provides a reproducible runtime environment for command-line processing.
+
 ## Current Features:
 **Sentinel-2**
 - [X] Sentinel-2 imagery reading
@@ -51,13 +62,12 @@ The project currently combines multispectral Sentinel-2 processing with classica
 - [X] Processing results export
 - [X] Batch processing of multiple Sentinel-2 products
 - [X] Visualization
-- [X] Unit Tests
-- [X] Command-Line Interface (CLI)
 
 **UAV RGB**
 - [X] RGB UAV imagery reading
 - [X] RGB vegetation indices
 - [X] Vegetation segmentation
+- [X] Vegetation fraction estimation
 - [X] Crop-row orientation estimation
 - [X] Crop-row detection
 - [X] Crop-row boundary detection
@@ -66,12 +76,22 @@ The project currently combines multispectral Sentinel-2 processing with classica
 - [X] Georeferenced UAV Image support
 - [X] Experimental plot metadata integration
 - [X] UAV RGB workflow demonstration
-
-**Software quality**
-- [X] Visualization utilities
-- [X] Unit tests
-- [X] Code quality checks with Ruff
+- [x] Georeferenced raster metadata validation
+- [x] GeoTIFF raster tiling with geospatial metadata preservation
+- [x] Reconstruction of raster mosaics from tiles
+
+**Software quality & distribution**
+
+- [X] Unit and integration tests with pytest
+- [X] Automated code quality checks with Ruff
+- [X] Automated CI with GitHub Actions
+- [X] Test coverage monitoring with pytest-cov and Codecov
+- [X] Python package build and distribution through PyPI
+- [X] Docker containerization
+- [X] Automated Docker image publishing through GitHub Actions
+- [X] Dependency update management with Dependabot
 - [X] Reproducible toy datasets
+- [X] Command-line interface (CLI)
 
 ## Case studies
 
@@ -676,6 +696,127 @@ opencropphenotyping batch-uav \
 
 The CLI follows the same processing pipeline as the Python API and can therefore be used for reproducible processing of individual UAV datasets or complete directories of datasets.
 
+## Georeferenced raster support and tiling
+
+OpenCropPhenotyping provides utilities to validate georeferenced raster metadata, split GeoTIFF rasters into smaller tiles, and reconstruct a raster mosaic from these tiles.
+
+These utilities support the handling of large UAV orthomosaics while preserving their spatial reference information.
+
+### Available functions
+
+- `validate_georeferenced_raster()`: checks the consistency of raster georeferencing metadata.
+- `tile_raster()`: divides a raster into smaller tiles while preserving their spatial metadata.
+- `merge_raster_tiles()`: reconstructs a raster mosaic from georeferenced tiles.
+
+The functions operate on existing georeferenced rasters. They do not generate orthomosaics from raw UAV photographs, which requires a dedicated photogrammetry workflow.
+
+## Docker
+
+OpenCropPhenotyping is available as a Docker image, providing a reproducible Linux environment with the package and its dependencies pre-installed.
+
+Docker images are published to GitHub Container Registry (GHCR) when a new version is released.
+
+### Requirements
+
+- [Docker Desktop](https://www.docker.com/products/docker-desktop/) or Docker Engine
+- UAV or Sentinel-2 input data (toy datasets are included in the image)
+- A published OpenCropPhenotyping Docker image
+
+### Download the image
+
+Pull a specific released version:
+
+```bash
+docker pull ghcr.io/aurelien-peter/opencropphenotyping:VERSION
+```
+
+Replace `VERSION` with an available release tag, such as `0.2.2`.
+
+### Check the installation
+
+Display the available CLI commands:
+
+```bash
+docker run --rm ghcr.io/aurelien-peter/opencropphenotyping:VERSION --help
+```
+
+Available commands include:
+
+- `process`: process a single Sentinel-2 product.
+- `batch`: process multiple Sentinel-2 products.
+- `segment-uav`: process a single UAV RGB dataset.
+- `batch-uav`: process multiple UAV datasets.
+
+### Process Sentinel-2 imagery
+
+Input data and output directories are provided through a bind mount. In the following example, a local `data` directory is mounted at `/data` inside the container.
+
+```bash
+docker run --rm \
+  --mount type=bind,source=/absolute/path/to/data,target=/data \
+  ghcr.io/aurelien-peter/opencropphenotyping:VERSION \
+  process \
+  --input-dir /data/raw/sentinel_product \
+  --output-dir /data/results \
+  --processed-dir /data/processed \
+  --indices ndvi \
+  --ndvi-threshold 0.3 \
+  --resolution 10
+```
+
+Replace `/absolute/path/to/data` with the absolute path to the local directory containing your data.
+
+The input path must point to a Sentinel-2 product supported by the pipeline. Results written under `/data` are available in the corresponding local directory after the container exits.
+
+### Process UAV RGB imagery
+
+A single UAV dataset requires an RGB image, its georeferenced GeoTIFF and a GeoPackage describing the experimental plot.
+
+```bash
+docker run --rm \
+  --mount type=bind,source=/absolute/path/to/data,target=/data \
+  ghcr.io/aurelien-peter/opencropphenotyping:VERSION \
+  segment-uav \
+  --image /data/uav/image.png \
+  --geotiff /data/uav/image_georef.tif \
+  --geopackage /data/uav/plots.gpkg \
+  --output-dir /data/results/uav
+```
+
+To process several UAV datasets, use the `batch-uav` command. Each dataset must be stored in its own subdirectory containing the required input files.
+
+```bash
+docker run --rm \
+  --mount type=bind,source=/absolute/path/to/data,target=/data \
+  ghcr.io/aurelien-peter/opencropphenotyping:VERSION \
+  batch-uav \
+  --input-dir /data/uav \
+  --output-dir /data/results/uav
+```
+
+### Build the image locally
+
+To build the Docker image from the source code instead of pulling a published release, clone the repository and run:
+
+```bash
+git clone https://github.com/Aurelien-Peter/OpenCropPhenotyping.git
+cd OpenCropPhenotyping
+
+docker build -t opencropphenotyping:local .
+```
+
+Check the local image:
+
+```bash
+docker run --rm opencropphenotyping:local --help
+```
+
+### Notes
+
+- The container is designed for command-line processing jobs. It starts, executes the requested command and exits when the job is complete.
+- Use a bind mount for both input data and output directories to keep files accessible outside the container.
+
+
 ## Contributing 
 
 Contributions are welcome. If you have ideas for improvements, bug fixes or new features, feel free to open an issue or submit a pull request.

{opencropphenotyping-0.2.0 → opencropphenotyping-0.3.0}/README.md

@@ -2,15 +2,23 @@
 
 [![CI](https://github.com/Aurelien-Peter/OpenCropPhenotyping/actions/workflows/ci.yml/badge.svg)](https://github.com/Aurelien-Peter/OpenCropPhenotyping/actions/workflows/ci.yml)
 [![codecov](https://codecov.io/github/Aurelien-Peter/OpenCropPhenotyping/graph/badge.svg?token=7SAC1M1VY5)](https://codecov.io/github/Aurelien-Peter/OpenCropPhenotyping)
+[![PyPI Version](https://img.shields.io/pypi/v/opencropphenotyping)](https://pypi.org/project/opencropphenotyping/)
+[![Python versions](https://img.shields.io/pypi/pyversions/opencropphenotyping)](https://pypi.org/project/opencropphenotyping/)
+[![Docker publish](https://github.com/Aurelien-Peter/OpenCropPhenotyping/actions/workflows/publish-docker.yml/badge.svg)](https://github.com/Aurelien-Peter/OpenCropPhenotyping/actions/workflows/publish_docker.yml)
 
 **An open-source Python framework for high-throughput crop phenotyping using drone and satellite imagery.**
 
 ## About the project
 
 OpenCropPhenotyping is an open-source project dedicated to reproducible high-throughput crop phenotyping using remote sensing data.
+
 The objective is to provide reproducible workflows to extract agronomic traits from UAV and Sentinel-2 imagery, from raw image preprocessing to trait extraction, statistical analysis and visualization. 
+
+
 The project currently combines multispectral Sentinel-2 processing with classical RGB UAV image analysis. The first UAV developments focus on crop-row detection, vegetation segmentation and plant-position analysis, with the objective of progressively extending the framework toward more advanced high-throughput phenotyping workflows.
 
+The project includes automated testing, continuous integration, code coverage monitoring, and package distribution through PyPI. Docker support provides a reproducible runtime environment for command-line processing.
+
 ## Current Features:
 **Sentinel-2**
 - [X] Sentinel-2 imagery reading
@@ -27,13 +35,12 @@ The project currently combines multispectral Sentinel-2 processing with classica
 - [X] Processing results export
 - [X] Batch processing of multiple Sentinel-2 products
 - [X] Visualization
-- [X] Unit Tests
-- [X] Command-Line Interface (CLI)
 
 **UAV RGB**
 - [X] RGB UAV imagery reading
 - [X] RGB vegetation indices
 - [X] Vegetation segmentation
+- [X] Vegetation fraction estimation
 - [X] Crop-row orientation estimation
 - [X] Crop-row detection
 - [X] Crop-row boundary detection
@@ -42,12 +49,22 @@ The project currently combines multispectral Sentinel-2 processing with classica
 - [X] Georeferenced UAV Image support
 - [X] Experimental plot metadata integration
 - [X] UAV RGB workflow demonstration
-
-**Software quality**
-- [X] Visualization utilities
-- [X] Unit tests
-- [X] Code quality checks with Ruff
+- [x] Georeferenced raster metadata validation
+- [x] GeoTIFF raster tiling with geospatial metadata preservation
+- [x] Reconstruction of raster mosaics from tiles
+
+**Software quality & distribution**
+
+- [X] Unit and integration tests with pytest
+- [X] Automated code quality checks with Ruff
+- [X] Automated CI with GitHub Actions
+- [X] Test coverage monitoring with pytest-cov and Codecov
+- [X] Python package build and distribution through PyPI
+- [X] Docker containerization
+- [X] Automated Docker image publishing through GitHub Actions
+- [X] Dependency update management with Dependabot
 - [X] Reproducible toy datasets
+- [X] Command-line interface (CLI)
 
 ## Case studies
 
@@ -652,6 +669,127 @@ opencropphenotyping batch-uav \
 
 The CLI follows the same processing pipeline as the Python API and can therefore be used for reproducible processing of individual UAV datasets or complete directories of datasets.
 
+## Georeferenced raster support and tiling
+
+OpenCropPhenotyping provides utilities to validate georeferenced raster metadata, split GeoTIFF rasters into smaller tiles, and reconstruct a raster mosaic from these tiles.
+
+These utilities support the handling of large UAV orthomosaics while preserving their spatial reference information.
+
+### Available functions
+
+- `validate_georeferenced_raster()`: checks the consistency of raster georeferencing metadata.
+- `tile_raster()`: divides a raster into smaller tiles while preserving their spatial metadata.
+- `merge_raster_tiles()`: reconstructs a raster mosaic from georeferenced tiles.
+
+The functions operate on existing georeferenced rasters. They do not generate orthomosaics from raw UAV photographs, which requires a dedicated photogrammetry workflow.
+
+## Docker
+
+OpenCropPhenotyping is available as a Docker image, providing a reproducible Linux environment with the package and its dependencies pre-installed.
+
+Docker images are published to GitHub Container Registry (GHCR) when a new version is released.
+
+### Requirements
+
+- [Docker Desktop](https://www.docker.com/products/docker-desktop/) or Docker Engine
+- UAV or Sentinel-2 input data (toy datasets are included in the image)
+- A published OpenCropPhenotyping Docker image
+
+### Download the image
+
+Pull a specific released version:
+
+```bash
+docker pull ghcr.io/aurelien-peter/opencropphenotyping:VERSION
+```
+
+Replace `VERSION` with an available release tag, such as `0.2.2`.
+
+### Check the installation
+
+Display the available CLI commands:
+
+```bash
+docker run --rm ghcr.io/aurelien-peter/opencropphenotyping:VERSION --help
+```
+
+Available commands include:
+
+- `process`: process a single Sentinel-2 product.
+- `batch`: process multiple Sentinel-2 products.
+- `segment-uav`: process a single UAV RGB dataset.
+- `batch-uav`: process multiple UAV datasets.
+
+### Process Sentinel-2 imagery
+
+Input data and output directories are provided through a bind mount. In the following example, a local `data` directory is mounted at `/data` inside the container.
+
+```bash
+docker run --rm \
+  --mount type=bind,source=/absolute/path/to/data,target=/data \
+  ghcr.io/aurelien-peter/opencropphenotyping:VERSION \
+  process \
+  --input-dir /data/raw/sentinel_product \
+  --output-dir /data/results \
+  --processed-dir /data/processed \
+  --indices ndvi \
+  --ndvi-threshold 0.3 \
+  --resolution 10
+```
+
+Replace `/absolute/path/to/data` with the absolute path to the local directory containing your data.
+
+The input path must point to a Sentinel-2 product supported by the pipeline. Results written under `/data` are available in the corresponding local directory after the container exits.
+
+### Process UAV RGB imagery
+
+A single UAV dataset requires an RGB image, its georeferenced GeoTIFF and a GeoPackage describing the experimental plot.
+
+```bash
+docker run --rm \
+  --mount type=bind,source=/absolute/path/to/data,target=/data \
+  ghcr.io/aurelien-peter/opencropphenotyping:VERSION \
+  segment-uav \
+  --image /data/uav/image.png \
+  --geotiff /data/uav/image_georef.tif \
+  --geopackage /data/uav/plots.gpkg \
+  --output-dir /data/results/uav
+```
+
+To process several UAV datasets, use the `batch-uav` command. Each dataset must be stored in its own subdirectory containing the required input files.
+
+```bash
+docker run --rm \
+  --mount type=bind,source=/absolute/path/to/data,target=/data \
+  ghcr.io/aurelien-peter/opencropphenotyping:VERSION \
+  batch-uav \
+  --input-dir /data/uav \
+  --output-dir /data/results/uav
+```
+
+### Build the image locally
+
+To build the Docker image from the source code instead of pulling a published release, clone the repository and run:
+
+```bash
+git clone https://github.com/Aurelien-Peter/OpenCropPhenotyping.git
+cd OpenCropPhenotyping
+
+docker build -t opencropphenotyping:local .
+```
+
+Check the local image:
+
+```bash
+docker run --rm opencropphenotyping:local --help
+```
+
+### Notes
+
+- The container is designed for command-line processing jobs. It starts, executes the requested command and exits when the job is complete.
+- Use a bind mount for both input data and output directories to keep files accessible outside the container.
+
+
 ## Contributing 
 
 Contributions are welcome. If you have ideas for improvements, bug fixes or new features, feel free to open an issue or submit a pull request.

{opencropphenotyping-0.2.0 → opencropphenotyping-0.3.0}/pyproject.toml

@@ -5,13 +5,18 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "opencropphenotyping"
-version = "0.2.0"
+version = "0.3.0"
 description = "Open-source tools for crop phenotyping using multispectral geospatial imagery"
 authors = [
     {name = "Aurélien Peter"}
 ]
 readme = "README.md"
 requires-python = ">=3.11"
+classifiers = [
+    "Programming Language :: Python :: 3",
+    "Programming Language :: Python :: 3.11",
+    "Operating System :: OS Independent",
+]
 
 dependencies = [
     "numpy",

{opencropphenotyping-0.2.0 → opencropphenotyping-0.3.0}/src/opencropphenotyping/indices.py

@@ -144,7 +144,166 @@ def compute_savi(red_band: np.ndarray, nir_band: np.ndarray, L_factor : float =
 
     return savi.astype(np.float32)
 
+def compute_exg(
+    rgb: tuple[np.ndarray, np.ndarray, np.ndarray],
+) -> np.ndarray:
+    """Compute the Excess Green Index (ExG).
+
+    Parameters
+    ----------
+    rgb : tuple[np.ndarray, np.ndarray, np.ndarray]
+        RGB image channels as (red, green, blue).
+
+    Returns
+    -------
+    np.ndarray
+        ExG image with dtype float32.
+    """
+    red, green, blue = (band.astype(np.float32) for band in rgb)
+
+    return 2 * green - red - blue
+
+def compute_exr(
+    rgb: tuple[np.ndarray, np.ndarray, np.ndarray],
+) -> np.ndarray:
+    """Compute the Excess Red Index (ExR).
+
+    Parameters
+    ----------
+    rgb : tuple[np.ndarray, np.ndarray, np.ndarray]
+        RGB image channels as (red, green, blue).
+
+    Returns
+    -------
+    np.ndarray
+        ExR image with dtype float32.
+    """
+    red, green, _ = (band.astype(np.float32) for band in rgb)
+
+    return 1.4 * red - green
+
+
+def compute_exgr(
+    rgb: tuple[np.ndarray, np.ndarray, np.ndarray],
+) -> np.ndarray:
+    """Compute the Excess Green minus Excess Red Index (ExGR).
+
+    Parameters
+    ----------
+    rgb : tuple[np.ndarray, np.ndarray, np.ndarray]
+        RGB image channels as (red, green, blue).
+
+    Returns
+    -------
+    np.ndarray
+        ExGR image with dtype float32.
+    """
+    exg = compute_exg(rgb)
+    exr = compute_exr(rgb)
+
+    return exg - exr
+
+
+def compute_vari(
+    rgb: tuple[np.ndarray, np.ndarray, np.ndarray],
+) -> np.ndarray:
+    """Compute the Visible Atmospherically Resistant Index (VARI).
+
+    Parameters
+    ----------
+    rgb : tuple[np.ndarray, np.ndarray, np.ndarray]
+        RGB image channels as (red, green, blue).
+
+    Returns
+    -------
+    np.ndarray
+        VARI image with dtype float32.
+    """
+    red, green, blue = (band.astype(np.float32) for band in rgb)
+
+    numerator = green - red
+    denominator = green + red - blue
+
+    result = np.full(numerator.shape, np.nan, dtype=np.float32)
+
+    np.divide(
+        numerator,
+        denominator,
+        out=result,
+        where=denominator != 0,
+    )
+
+    return result
+
+
+def compute_gli(
+    rgb: tuple[np.ndarray, np.ndarray, np.ndarray],
+) -> np.ndarray:
+    """Compute the Green Leaf Index (GLI).
+
+    Parameters
+    ----------
+    rgb : tuple[np.ndarray, np.ndarray, np.ndarray]
+        RGB image channels as (red, green, blue).
+
+    Returns
+    -------
+    np.ndarray
+        GLI image with dtype float32.
+    """
+    red, green, blue = (band.astype(np.float32) for band in rgb)
+
+    numerator = 2 * green - red - blue
+    denominator = 2 * green + red + blue
+
+    result = np.full(numerator.shape, np.nan, dtype=np.float32)
+
+    np.divide(
+        numerator,
+        denominator,
+        out=result,
+        where=denominator != 0,
+    )
+
+    return result
+
+
+def compute_rgbvi(
+    rgb: tuple[np.ndarray, np.ndarray, np.ndarray],
+) -> np.ndarray:
+    """Compute the Red Green Blue Vegetation Index (RGBVI).
+
+    Parameters
+    ----------
+    rgb : tuple[np.ndarray, np.ndarray, np.ndarray]
+        RGB image channels as (red, green, blue).
+
+    Returns
+    -------
+    np.ndarray
+        RGBVI image with dtype float32.
+    """
+    red, green, blue = (band.astype(np.float32) for band in rgb)
+
+    green_squared = green**2
+    red_blue = red * blue
+
+    numerator = green_squared - red_blue
+    denominator = green_squared + red_blue
+
+    result = np.full(numerator.shape, np.nan, dtype=np.float32)
+
+    np.divide(
+        numerator,
+        denominator,
+        out=result,
+        where=denominator != 0,
+    )
+
+    return result
+
 VEGETATION_INDICES = {
+    # Satellite
     "ndvi": {
         "function": compute_ndvi,
         "bands": ["red_band", "nir_band"],
@@ -161,6 +320,37 @@ VEGETATION_INDICES = {
         "function": compute_gndvi,
         "bands": ["nir_band", "green_band"],
     },
+    # UAV RGB
+    "exg": {
+        "bands": ["red_band", "green_band", "blue_band"],
+        "function": compute_exg,
+        "input_type": "rgb",
+    },
+    "exr": {
+        "bands": ["red_band", "green_band", "blue_band"],
+        "function": compute_exr,
+        "input_type": "rgb",
+    },
+    "exgr": {
+        "bands": ["red_band", "green_band", "blue_band"],
+        "function": compute_exgr,
+        "input_type": "rgb",
+    },
+    "vari": {
+        "bands": ["red_band", "green_band", "blue_band"],
+        "function": compute_vari,
+        "input_type": "rgb",
+    },
+    "gli": {
+        "bands": ["red_band", "green_band", "blue_band"],
+        "function": compute_gli,
+        "input_type": "rgb",
+    },
+    "rgbvi": {
+        "bands": ["red_band", "green_band", "blue_band"],
+        "function": compute_rgbvi,
+        "input_type": "rgb",
+    },
 }
 
 def compute_indexes(
@@ -168,6 +358,7 @@ def compute_indexes(
         nir_band: np.ndarray | None = None,
         green_band: np.ndarray | None = None,
         red_edge_band: np.ndarray | None = None,
+        blue_band: np.ndarray | None = None,
         indices: list[str] | None = None,
     ) -> dict[str, np.ndarray]:
     """Compute vegetation indices from available spectral bands.
@@ -182,6 +373,8 @@ def compute_indexes(
         Green spectral band.
     red_edge_band : np.ndarray or None, optional
         Red-edge spectral band.
+    blue_band : np.ndarray or None, optional
+        Blue spectral band.
     indices : list[str] or None, optional
         Names of the vegetation indices to compute. If None, all
         available vegetation indices are requested.
@@ -201,13 +394,14 @@ def compute_indexes(
     results = {}
 
     if indices is None:
-        indices = ["ndvi", "savi", "ndre", "gndvi"]
+        indices = list(VEGETATION_INDICES)
 
     available_bands = {
         "red_band": red_band,
         "nir_band": nir_band,
         "green_band": green_band,
         "red_edge_band": red_edge_band,
+        "blue_band": blue_band,
     }
 
     for index in indices:
@@ -227,27 +421,9 @@ def compute_indexes(
                 for band in required_bands
             ]
 
-            results[index] = compute_function(*band_arrays)
+            if index_config.get("input_type") == "rgb":
+                results[index] = compute_function(tuple(band_arrays))
+            else:
+                results[index] = compute_function(*band_arrays)
             
-    return results
-    
-def compute_exg(
-    rgb: tuple[np.ndarray, np.ndarray, np.ndarray],
-) -> np.ndarray:
-    """Compute the Excess Green Index (ExG).
-
-    Parameters
-    ----------
-    rgb : tuple[np.ndarray, np.ndarray, np.ndarray]
-        RGB image channels as (red, green, blue).
-
-    Returns
-    -------
-    np.ndarray
-        ExG image with dtype float32.
-    """
-    red = rgb[0].astype(np.float32)
-    green = rgb[1].astype(np.float32)
-    blue = rgb[2].astype(np.float32)
-
-    return 2 * green - red - blue
+    return results