loicdtx / micamac

micasense orthomosaic using micmac

MicaMAC

multispectral ortho-mosaics for micasense redEdge cameras using MICMAC

• Micasense RedEdge camera
• MICMAC

MicaMAC is a set of wrappers to facilitate the generation of ortho-mosaics out of micasense RedEdge data; it uses the micasense python package, MICMAC and a bit of python.

Processing micasense RedEdge data does not work out of the box with free/open-source photogrammetry software tools like MICMAC mainly for 3 reasons:

• The individual bands of the raw captures are not aligned with each others
• There are five bands while most software are optimized for 3 bands (RGB) images.
• Many software do not support tiff file format and/or data in int16. Again, optimized for "classic" photographs (3 bands jpg image in int8). MICMAC is the exception there as it's perfectly happy using tiff in int16.

Usage

There are two command lines:

• align_images.py performs bands alignent, optional altitude and AOI filtering, and optional conversion to surface reflectance.
• run_micmac.py runs a more or less standard micmac workflow on the results of the align_images.py command, resulting in the generation of ortho-mosaic, DEM and dense point cloud.

Both command lines have a detailed manual that can be accessed by running the command with the --help flag.

Installation

1. Install MICMAC (see installation guide )
2. Clone the repos:

git clone https://github.com/loicdtx/micamac.git

3. Install python dependencies and package (preferably inside a python3 virtualenv)

cd micamac
pip install -r requirements.txt
pip install -e .

If it fails because of GDAL or rasterio, try running the commands below.

pip install numpy
pip install GDAL==$(gdal-config --version) --global-option=build_ext --global-option="-I/usr/include/gdal"
pip install -r requirements.txt
pip install -e .

4. Optionally if you have flights without proper panel captures and DLS but still want some sort of conversion to reflectance/normalization among flights (using 6S radiative transfer modeling), you can use the irraiance-modeling branch. For that you must install Py6S (and 6S).