Welcome to plaknit 🧶¶
Processing Large-Scale PlanetScope Data
- Free software: MIT License
- Documentation: https://dzfinch.github.io/plaknit
Why plaknit exists¶
PlanetScope Scene (PSS) data are reveared for its quality and distinct ability to balance spatial and temporal resolution in Earth Observation data. While PSS has proven itself a valuable asset in monitoring small-scale areas, the literature has pointed out the shortcomings when creating a single image from individual tiles (Frazier & Hemingway, 2021).
plaknit bundles the workflow I use to operationalize large-area mosaics so
you can run the same process locally or in an HPC environment. The goal is to
spend time answering big questions, not making a big mess of your data.
Features¶
- Run everything from a single CLI (
plaknit) that works cross-platform (plan, order, mosaic). - Plan PSScene acquisitions per month (
plaknit plan), auto-simplify ROIs to Planet’s 1,500-vertex limit, and submit resilient Planet Orders (plaknit order). - Build seamless mosaics (
plaknit mosaic) with pre-tuned Orfeo Toolbox parameters and RAM hints after masking PlanetScope tiles against their UDM rasters using efficient GDAL workflows. - Train and apply Random Forest classifiers with optional MRF smoothing on multi-band stacks directly from CLI with
plaknit classify trainandplaknit classify predict(Bayesian smoothing option in progress).
Need to run on HPC?¶
See Running plaknit on HPC with Singularity/Apptainer for a copy-pasteable recipe that uses persistent virtual environments and SLURM batch jobs inside containerized OTB builds.