Implement SAR texture measures based on co-occurence matrices
gilbertocamara opened this issue · comments
Describe the new API function requested
Reccent papers on deforestation alerts, as for example "How textural features can improve SAR-based tropical forest disturbance mapping" indicate that some of the Haralick texture metrics based on co-occurence matrix can improve their accuracy.
For this reason, we should consider a new function sits_sar_texture()
that implements the texture measures described in Table 2 of the above paper.
Associated sits API function
sits_sar_texture(cube, measure, output_dir, multicores, memzise)
where:
cube is a SAR image data cube and measure
is one of grey-level co-occurence matrices (GLCM) metrics.
@gilbertocamara you may be interested in https://github.com/ailich/GLCMTextures by @ailich
Hi @Nowosad many thanks for the very useful tip!
@gilbertocamara you are welcome.
I think it would be great to have one high quality and comprehensive package for GLCM textures than a few ones only having some measures...