Use GLCM to add image texture in SNAP

Microwave Toolbox
#1

I wish to perform classification on my 1SDV GRD product. I have performed the following:
Calibration
Speckle Filtering
Terrain correction

Can anyone tell me how to used GLCM matrix in SNAP to add image texture in images so that they can be used for classification?

Thank You.

1 Like
#2

if you want to derive image texture, I would recommend to skip speckle filtering because it destroys most of the texture.

To use it for classification, simply run the GLCM module on your calibrated (and terrain corrected) data and use the result as the input for the classification module.

3 Likes
#3

Hi Everyone! Sorry for post here, but i have slighly the same question.

I wish to perform classification on my S1 product
But i dont understand why we performed it on GRD and not on SLC image. I understand we need the raw pixels for better classification.

Therefore, on S1 GRD, i do
Orbit
Calibration
Terrain correction

Now, when i do this, maybe i am doing some changes to the original image too, or im wrong ?

Thank You!
Guido

#4

both SLC and GRD contain backscatter intensity. If you don’t need the phase information for classification (e.g. for the computation if interferometric coherence), there is little reason to use SLC. The only difference is that SLC has a resolution of 5x20m which is multi-looked to 10x10m in GRD format. This is more convenient an requires significantly less disk space.
Theoretically, you could use SLC, apply Orbit files and TOPS Split, then deburst and calculate the image textures on the SLC intensities. But the differences are mostly minor.
So using GRD, applying Orbit files, then calibrate to Sigma0, then apply Terrain Correction is fully sufficient.