Dear all,
I am trying to classify urban areas using Landsat TM data, I have tried to get the image texture using GLCM and the result is somehow good, but I have seen many other literatures which exploited Multivariate Image texture and had better result compare to GLCM method.
GLCM: uses one band as input data
Multivariate Img text: uses multi-band as input data
here is a paper regarding Multivariate Img Text : https://pdfs.semanticscholar.org/e5b4/9480f331eaf33de24f48b2304060438f2333.pdf
I appreciate your help.
regards,
Salim