Analyze multiband raster using PCA to find the most unique and informative band combinations
PCA Band Selector performs Principal Component Analysis (PCA) on any multiband raster (Landsat, Sentinel, AVIRIS, etc.) to identify which bands carry the most unique spectral information. It recommends optimal 2, 3, and 4-band combinations for classification and visualization by maximizing information content and minimizing inter-band redundancy.
Features:
- Explained variance chart per Principal Component
- PC Loadings matrix (color coded)
- Band uniqueness scores ranked
- Best 2/3/4-band combination recommendations
- Inter-band correlation matrix
- Auto memory management for large images
- Export results as TXT report or CSV
- Compatible with QGIS 3.x and QGIS 4.x (Qt5 and Qt6)
Plugin Tags