Sentera Radiometric Corrections

Tool to perform various corrections on imagery from supported sensors, including Sentera 6X, Sentera Double 4K and DJI Mavic 3 Multispectral.

Installation

To install the Sentera Radiometric Corrections app, download the Sentera Radiometric Corrections GUI.zip file from the latest release on the homepage. After downloading, extract the files and run the executable file.

The Sentera Radiometric Corrections tool will perform corrections on multispectral drone imagery as well as recflectance corrections when a supported Sentera calibration panel is used.

Required Inputs

• Input Path: Folder path containing multispectral imagery. Provide an individual folder of images or a folder containing images from multiple sensors.
• Output Path: Folder path to save the correct images to.

Optional Inputs/Parameters

• Reflectance Correction: Select reflectance correction to perform reflectance corrections using a Sentera calibration panel. If calibration panel imagery is not detected in the input images, this option will be disabled.
• ILS Correction: Select ILS correction to perform lighting corrections using a light sensor. If ILS values are not detected in the input images this option will be disabled.

**Note: If Reflectance and ILS corrections are disabled, images will be corrected for exposure only

Advanced Inputs/Parameters

• ExifTool Path: Optionally provide an updated/custom version of exiftool. If not provided the default version is included in the tool.
• Calibration ID: For Sentera sensors, calibration images are tagged with "CAL". Optionally provide a different tag if the image names have been modified. For non-Sentera sensors all images are scanned for calibration panels.
• Use All Panel Sets(6X): By default only the best calibration set will be used to perform corrections for Sentera sensors. Select this option to perform corrections using all available calibration panel captures.
• Delete/Overwrite Original Images: Select this option and set the input and output paths to match to overwrite the original images with the corrected images. Not reccomended as original images cannot be restored.
• Output as UInt16(0-65535): By default corrected images are formated in Float32 with values in reflectance ranging from 0-1. Selecting this option will output the corrected images in Uint16 with values ranging from 0-65535. This may be required for certain photogrametry software and corrected images will have a much smaller file size.

Installation instructions for usage/development in Python

Windows (Conda)

1. Download Miniconda for Python3.7

2. Open Anaconda Prompt and clone py-radiometric-corrections with

    >> git clone https://github.com/SenteraLLC/py-radiometric-corrections.git
   

3. Open Anaconda Prompt and navigate to py-radiometric-corrections/. Run:

    >> conda env create -f environment.yml
    >> conda activate imgcorrect-venv
    >> pip install .
   

4. This creates an imgcorrect-venv environment that all scripts should be run in and installs the analyticstest library for the scripts to reference.

If no errors appear, the imgcorrect library should be installed correctly.

Usage

The imagery correction in this repository can be used via:

• Importing the various library functions defined in the package
• Running the pre-defined scripts with a Python installation of version 3.6 or above
• Running the standalone executable on the command line

Open Anaconda Prompt and navigate to py-radiometric-corrections/. Run:

    >> conda activate imgcorrect-venv

Get a list of required arguments:

>> python scripts\correct_images.py -h 

Run corrections:

>> python scripts\correct_images.py "input_path" -o "output_path" **Additional optional commands

Optional commands:

--calibration_id "CALIBRATION_ID", -c CALIBRATION_ID

• Identifier in the name of the image that denotes it is from the calibration set. If not specified, defaults to "CAL".

--output_path OUTPUT_PATH, -o OUTPUT_PATH

• Path to output folder at which the corrected images will be stored. If not supplied, corrected images will be placed into the input directory.

--no_ils_correct, -i

• If selected, ILS correction will not be applied to the images.

--no_reflectance_correct, -r

• If selected, reflectance correction will not be applied to the images.

--all_panels, -a

• If selected, reflectance correction will be applied using all panel images.

--delete_original, -d

• Overwrite original 12-bit images with the corrected versions. If selected, corrected images are renamed to their original names. If not, an extension is added.

--exiftool_path EXIFTOOL_PATH, -e EXIFTOOL_PATH

• Path to ExifTool executable. ExifTool is required for the conversion; if not passed, the script will use a bundled ExifTool executable.

--uint16_output, -u

• If selected, scale of output values will be adjusted to 0-65535 and dtype will be changed to uint16.

Building the Executable

In a Windows 10/11 x64 environment, rebuild the executable with pyinstaller using this command:

	>> pyinstaller correct_images_onefile.spec