Classifier


This module implements methods to detect patterns in image regions. Commonly, classification algorithms are divided by the level of classification (pixel or region), and by the interaction of the user (supervised or unsupervised). Pixel-based algorithms classify individual pixels according to their resemblance to a specific pattern.  Region-based algorithms use regions from segmented images, and classify each region to a specific pattern. Supervised methods uses a predefined typology, given by the user, who supplies samples of each pattern. Unsupervised methods detect an unknown number of patterns, according to their own method.

The available methods in TerraLib are:



ISOSeg

This is an unsupervised and region-based classification algorithm.

Input:

Expectation-Maximization - EM

This is an unsupervised and pixel-based classification algorithm. Expectation-Maximization (EM) works iteratively by applying two steps: the E-step (Expectation) and the M-step (Maximization). 

Input:

K-Means

This is an unsupervised and pixel-based classification algorithm.

Input:

Spectral Angle Mapper - SAM

This is an supervised classification algorithm.

Input:

Maximum a Posteriori Probability - MAP

This is an supervised classification algorithm.

Input:

Maximum Likelihood/ICM (a.k.a MaxVer/ICM)

This is a supervised classification algorithm.

Input:

Parameters:

References:


It is accessible through:

   Raster Processing > Classification... (list of all raster layers will be available) 

This wizard consists of the following steps:

    Wizard Page 1 - Selection of the layer to execute the operation (Layer Search)

  1. On the List of Layers select the raster layer to apply the operation. 

  2. Optionally use Filter By Name field giving part of the layer name to help find the layer in the list.

  3. Press Next  to go to next step or Cancel to close the dialog. 

    Wizard Page 2 - Classification parameters

  1. Select the type of classifier to be used.

  2. Select the bands to be used in the process.

  3. As described above, each classifier has a set of specific attributes. For supervised classifiers (SAM and MAP)  is necessary to use a component for the acquisition of samples.

  4. Press Next  to go to next step, Back to return to the previous wizard or Cancel to close the dialog.

    Wizard Page 3 - Output information

  1. Raster Info - First press and inform the folder where the resulting file will be saved.

  2. Name - inform the raster name.

  3. Extra Parameters - if there are some, see the details on how to inform then here.

  4. Press Finish to save the resulting contrasted raster or Back to go to the previous wizard page.

Hint: The resulting image will be added as a new layer at the TerraView project.