[general]
name=Geomaticape Tools
description=Landsat C2 L1/L2, Sentinel-2, CBERS-04A, MODIS 09/11/12/13 scale factor per product independently, PCA, spectral indices, supervised + unsupervised classification, band extraction, vector geoprocessing, zonal statistics and multi-point sampling
version=1.8
qgisMinimumVersion=3.40
qgisMaximumVersion=4.99
author=Geomatica Ambiental
email=nino@geomatica.pe
authorUrl=https://www.geomatica.pe/

about=GeomaticaPE is a QGIS plugin developed by Geomatica Ambiental to automate
    preprocessing and analysis of multispectral satellite imagery and to
    support common geoprocessing tasks.

    Tools organized into three groups:

    [Conversion]
    - RS Landsat C2 L1 (SR + LST + PAN, autodetects MSS/TM/ETM/OLI)
    - Landsat C2 L2 scale factor
    - Sentinel-2 L1A scale factor
    - Sentinel-2 L2A scale factor

    [Processing]
    - CBERS-04A Pansharpening Brovey 2m
    - Landsat Pansharpening 30m -> 15m (Weighted Brovey)
    - Multispectral PCA
    - Spectral indices (17 indices)
    - Extract bands from multispectral images:
      detects number of bands
    - Clip raster by study area:
    - Combine bands with names
    - Supervised classification and validation (DT/RF/NB/MLP/KNN).
    
    [Geoprocessing]
    - Create polygons from table (CSV/TXT/XLS/XLSX)
    - Raster zonal statistics (Excel/CSV)
    - Extract point values from multiple rasters

    COMPATIBILITY:
    - Qt5 (PyQt5) and Qt6 (PyQt6) via qgis.PyQt compatibility layer
    - QGIS 3.40 to 4.99

    REQUIREMENTS - Python modules bundled with QGIS (no installation needed):
    - numpy           (all tools)
    - GDAL / osgeo    (all tools)
    - matplotlib      (QGIS LTR 3.22+; supervised_classification, pca_satellite)

    REQUIREMENTS - External modules NOT bundled with QGIS (must install):
    - scikit-learn    REQUIRED by: Unsupervised Classification (KMeans,
                      MiniBatchKMeans, GaussianMixture, Birch, StandardScaler),
                      Supervised Classification (DecisionTree, RandomForest,
                      NaiveBayes, MLP, KNN, train_test_split, classification_report),
                      Multispectral PCA (PCA, StandardScaler, make_pipeline)
    - pandas          REQUIRED by: Multispectral PCA (variance table),
                      Create Polygons from Table (read CSV/TXT/XLS/XLSX)
    - openpyxl        REQUIRED by: Raster Zonal Statistics (export .xlsx),
                      Extract Point Values (export .xlsx),
                      Classification Report (export .xlsx with charts),
                      Create Polygons from Table (read .xlsx)

    Install all external dependencies (run once):
       python -m pip install --upgrade scikit-learn pandas openpyxl

    Developed and maintained by:
    Geomatica Ambiental - www.geomatica.pe

    Acknowledgments:
    - Point Sampling Tool by Borys Jurgiel (GPL v2+) - inspiration for the
      multi-raster point sampling tool. https://github.com/borysiasty/pointsamplingtool

icon=Icons/logo_geomatica.png

tracker=https://www.geomatica.pe/
repository=https://www.geomatica.pe/
homepage=https://www.geomatica.pe/
category=Raster
tags=landsat, c2 l1, c2 l2, mss, tm, etm, oli, tirs, dos1, surface reflectance, lst, pansharpening, weighted brovey, brovey, vrt pansharpened, clip raster, cutline, bbox, study area, clipping, sentinel, cbers, pca, ndvi, kmeans, isodata, gmm, birch, supervised classification, decision tree, random forest, extra trees, gradient boosting, naive bayes, mlp, knn, svm, lda, qda, logistic regression, kappa, confusion matrix, balanced accuracy, f1, cross validation, band extraction, band stacking, band combination, multiband, spectral indices, remote sensing, raster, zonal statistics, vector, polygon, csv, xlsx, point sampling
hasProcessingProvider=yes
server=False
experimental=False
deprecated=False

changelog=1.8
- [Processing] Spectral Signature: Landsat 5/7/8/9, Sentinel-2, ASTER L1T
  Extraction by class, mean/min/max, PNG chart, Excel export