We are attempting to generate a landscape classification of a large area using remote sensing data with a 20 x 20 m cell size and an area of 200,000 km2. We have 8 different rasters and want to perform cluster analysis on them.
Obviously, the datasets are massive and we are struggling to generate a classification that groups pixels into clusters.
Any advice on the best methods/approach for classifying this large data set would be most appreciated.
In my opinion the density based clustering algorithms are an appropriate choice for the large datasets.
https://desktop.arcgis.com/en/arcmap/10.3/tools/spatial-analyst-toolbox/iso-cluster-unsupervised-classification.htm
https://pro.arcgis.com/es/pro-app/help/analysis/image-analyst/classify.htm
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4699692/
http://www.academia.edu/download/602251/sia6.pdf
https://ieeexplore.ieee.org/abstract/document/6034316/
https://www.esri.com/arcgis-blog/products/spatial-analyst/analytics/whats-new-for-spatial-analyst-and-raster-analysis-in-arcgis-pro-2-3-2-4-and-desktop-10-7/
Article Land Cover and Land Use Classification with TWOPAC: towards ...
- Badges
- Science topic