Texturegui

Purpose

GUI for texture.

Synopsis

texturegui

texturegui(data)

Description

Main GUI for texture analysis.

Example

The 'textures' demo dataset contains 27 (128 pixel x 128 pixel) grayscale images of various composition. This dataset can be used to demonstrate the use of texture analysis via the 'texturegui' interface. In this example we'll load the data via the Image Manager, perform texture analysis, then export the texture data to the Analysis interface for further exploration.

• Open the Image Manager and load the Texture Photos from the Demo Data.
• From the Image Manager/Edit menu select Add All Images to Group.
• In the toolbar, click the Export to Textures button.

• The Texture Analysis window will open and images will be loaded.
• Select a Texture Method from the drop-down menu (lower left corner).
• Adjust any Method Settings in the panel below. NOTE: See the texture function for more information on texture method settings.
• Click Calculate. Note that some methods (semivar) will take significantly longer to calculate than others.
• The combined results will be plotted in the top axes and individual images along with resulting "spectra" plot are displayed below. Scroll down to see all images.

• The results of a particular method will be retained until the Method Settings are changed. This makes it easy to see results from two different methods without having to recalculate by simply changing the Method in the drop-down menu.
• The toolbar contains tools to help inspect and analyze results. The Magnify button will allow zooming, the Plot All Textures button will open combined spectra in a dedicated plot and the Export button will export combined data to the Analysis window.
• In this example, FFT is used with the spectrum direction of "columns" selected.
• Once a method and settings produces a desired result export the "spectra" to the Analysis interface for further exploration.
• Select PCA from the Methods menu and calculate a quick PCA model.
• If you open a Scores plot and look at the Q vs T2 plot you'll notice the more unique images are high Q like "Raffia", "Herringbone weave", and "Reptile skin".

See Also

autocor_img, detrend_img, fft_texture, semivar, svd_texture, texture