PCA Explorer
The PCA Explorer provides visualization, adjustment, clustering, selection, and export of principal-component data.
IDCubePro 2026 - PCA Explorer
Overview
The PCA Explorer provides visualization, adjustment, clustering, selection, and export of principal-component data.
The PCA Composite displays selected principal components as an RGB image.
Red channel - component assigned to the red display channel.
Green channel - component assigned to the green display channel.
Blue channel - component assigned to the blue display channel.
Use the sliders or numeric fields to select the PCA component assigned to each channel.
The Weight fields control the relative contribution of each PCA component to the RGB composite.
A value of 1 preserves the default contribution.
Contrast (%) adjusts the displayed intensity range of the PCA composite.
This changes visualization only and does not modify the PCA data.
Gamma adjusts the brightness response of the PCA composite.
Gamma affects visualization only.
Restores the default PCA composite settings: PC1 / PC2 / PC3, weights = 1, contrast = 0, and gamma = 1.
Displays the loading spectrum associated with each PCA component currently assigned to the red, green, and blue channels.
Shows the cumulative fraction of variance explained by the PCA components.
A rapid approach toward 1 indicates that most variance is represented by the first few components.
Clustering groups image pixels according to their position in PCA score space.
Select the desired number of clusters (2-25).
Select the distance metric used by K-medoids clustering.
Squared Euclidean is a useful general starting point.
Runs K-medoids clustering using the PCA score data.
The resulting classes are displayed as a cluster image and PCA scatter plot.
2D displays clusters using PC1 versus PC2.
3D displays clusters using PC1, PC2, and PC3.
Changing the view does not rerun clustering.
After clustering, use the class menu to display an individual cluster.
Selecting All Classes restores the complete clustering result.
Lasso provides manual selection directly in PCA score space.
1. Display the cluster scatter in 2D.
3. Draw a polygon around the desired PCA population.
4. Complete the polygon to create the selection.
The corresponding image pixels are highlighted in the Cluster Image.
Clears the current clustering result and PCA-space selection.
The PCA decomposition itself is not removed.
Exports image data corresponding to the currently selected class or Lasso region.
Exports the complementary image region: all valid pixels outside the current selection.
Exports the current selection as a binary spatial mask.
Exports the complete PCA score datacube.
Each image plane corresponds to one principal component.
1. Inspect the PCA composite.
2. Review the explained variance.
3. Select useful PCA components.
4. Adjust weights, contrast, or gamma if needed.
5. Choose the number of clusters and distance metric.
7. Inspect clusters in 2D or 3D.
8. Select a class or use Lasso for manual selection.
9. Export the selected region, inverse region, or mask as needed.