10 Jul 2020

Optical Thin Section Pore Space Characterization in HSV colorspace

Optical Thin Section Pore Space is segmented converting RGBA images in HSV colorspace by means of Python scripts.


In Optical Thin Sections, typically pore space is filled with blue epoxy resin. Segmenting the pore space in RGB is usually challenging; it is more intuitive in HSV colorspace.

The workflow converts the input RGBA image into an HSV colorspace, thresholding it with H,S,V range slides.

RemoveSmallSpot and RemoveSmallHoles is applied to clean up the segmentation.

Separate Pore Space is used to isolate each pore. Label Analysis is used to compute Length and Area in 2D.

Distribution Analysis is used to compute a binning histogram of Pore Size Distribution.

This Xtra contains input data and all intermediate steps, in addition to two python scripts:

  • conversion from RGB to HSV
  • HSV interactive threshold

A video shows how to use the modules on the given 2D image.

Download a version compatible with PerGeos 2020.1 to 2020.3 HERE.

Credit: U.S. Geological Survey, photo by Christopher Nye https://commons.wikimedia.org/wiki/File:Photomicrograph_of_rock_thin_section_Gabbroic_inclusion.jpg The workflow overview has been realized using a dataset from Stratum Reservoir