The surface roughness / localized standard deviation () is derived from a specified image or DTM / DEM. Image can also multi-channel, in that case roughness will be calculated channel wise.
where is the number of elements in the specified kernel.
Usage: roughness.py [-h] -i Input file path -o Output file path [-k [Kernel size ...]] [-p Padding mode] [-d Target device]
Calculate surface roughness / localized standard deviation
optional arguments:
-h, --help show this help message and exit
-i Input file path, --input Input file path
Specify input file
-o Output file path, --output Output file path
Specify output file
-k [Kernel size ...], --kernel [Kernel size ...]
Specify kernel size. Hint: int or int int
-p Padding mode, --padding Padding mode
Specify padding mode. Choices: 'zeros', 'reflect', 'replicate' or 'circular'. Default: 'replicate'
-d Target device, --device Target device
Specify target device. Default: 'cpu'