palisades 4.135.1

🧑🏽‍🚒 Post-Disaster Land Cover Classification.

🧑🏽‍🚒 palisades

🧑🏽‍🚒 Post-disaster land Cover classification using Semantic Segmentation on Maxar Open Data acquisitions.

pip install palisades

graph LR
    palisades_ingest_target["palisades<br>ingest -<br>target=&lt;target&gt; -<br>predict"]

    palisades_ingest_query["palisades<br>ingest -<br>&lt;query-object-name&gt; -<br>predict"]

    palisades_label["palisades<br>label<br>offset=&lt;offset&gt; -<br>&lt;query-object-name&gt;"]

    palisades_train["palisades<br>train -<br>&lt;query-object-name&gt; -<br>&lt;dataset-object-name&gt; -<br>&lt;model-object-name&gt;"]

    palisades_predict["palisades<br>predict - - -<br>&lt;model-object-name&gt;<br>&lt;datacube-id&gt;<br>&lt;prediction-object-name&gt;"]

    palisades_buildings_download_footprints["palisades<br>buildings<br>download_footprints -<br>&lt;input-object-name&gt; -<br>&lt;output-object-name&gt;"]

    palisades_buildings_analyze["palisades<br>buildings<br>analyze -<br>&lt;object-name&gt;"]

    target["🎯 target"]:::folder
    query_object["📂 query object"]:::folder
    datacube["🧊 datacube"]:::folder
    dataset_object["🏛️ dataset object"]:::folder
    model_object["🏛️ model object"]:::folder
    prediction_object["📂 prediction object"]:::folder

    query_object --> datacube

    target --> palisades_ingest_target
    palisades_ingest_target --> palisades_ingest_query
    palisades_ingest_target --> query_object

    query_object --> palisades_ingest_query
    palisades_ingest_query --> datacube
    palisades_ingest_query --> palisades_predict

    query_object --> palisades_label
    palisades_label --> datacube

    query_object --> palisades_train
    palisades_train --> dataset_object
    palisades_train --> model_object

    model_object --> palisades_predict
    datacube --> palisades_predict
    palisades_predict --> palisades_buildings_download_footprints
    palisades_predict --> palisades_buildings_analyze
    palisades_predict --> prediction_object

    prediction_object --> palisades_buildings_download_footprints
    palisades_buildings_download_footprints --> prediction_object

    prediction_object --> palisades_buildings_analyze
    palisades_buildings_analyze --> prediction_object

    classDef folder fill:#999,stroke:#333,stroke-width:2px;

palisades help

palisades \
	ingest \
	[~download,dryrun] \
	[target=<target> | <query-object-name>] \
	[~ingest | ~copy_template,dryrun,overwrite,scope=<scope>,upload] \
	[predict,count=<count>,~tag] \
	[device=<device>,profile=<profile>,upload] \
	[-|<model-object-name>] \
	[~download_footprints | country_code=<iso-code>,country_name=<country-name>,overwrite,source=<source>] \
	[~analyze | buffer=<buffer>,count=<count>] \
	[~submit | dryrun,to=<runner>]
 . ingest <target>.
   target: Brown-Mountain-Truck-Trail | Brown-Mountain-Truck-Trail-all | Brown-Mountain-Truck-Trail-test | Palisades-Maxar | Palisades-Maxar-test
   scope: all + metadata + raster + rgb + rgbx + <.jp2> + <.tif> + <.tiff>
      all: ALL files.
      metadata (default): any < 1 MB.
      raster: all raster.
      rgb: rgb.
      rgbx: rgb and what is needed to build rgb.
      <suffix>: any *<suffix>.
   device: cpu | cuda
   profile: FULL | DECENT | QUICK | DEBUG | VALIDATION
   country-name: for Microsoft, optional, overrides <iso-code>.
   iso-code: Country Alpha2 ISO code: https://en.wikipedia.org/wiki/List_of_ISO_3166_country_codes
      Canada: CA
      US: US
   source: microsoft | osm | google
   calls: https://github.com/microsoft/building-damage-assessment/blob/main/download_building_footprints.py
   buffer: in meters.
   runner: aws_batch | generic | local

palisades \
	label \
	[download,offset=<offset>] \
	[~download,dryrun,~QGIS,~rasterize,~sync,upload] \
	[.|<query-object-name>]
 . label <query-object-name>.

palisades \
	train \
	[dryrun,~download,review] \
	[.|<query-object-name>] \
	[count=<10000>,dryrun,upload] \
	[-|<dataset-object-name>] \
	[device=<device>,dryrun,profile=<profile>,upload,epochs=<5>] \
	[-|<model-object-name>]
 . train palisades.
   device: cpu | cuda
   profile: FULL | DECENT | QUICK | DEBUG | VALIDATION

palisades \
	predict \
	[~tag] \
	[~ingest | ~copy_template,dryrun,overwrite,scope=<scope>,upload] \
	[device=<device>,profile=<profile>,upload] \
	[-|<model-object-name>] \
	[.|<datacube-id>] \
	[-|<prediction-object-name>] \
	[~download_footprints | country_code=<iso-code>,country_name=<country-name>,overwrite,source=<source>] \
	[~analyze | buffer=<buffer>,count=<count>]
 . <datacube-id> -<model-object-name>-> <prediction-object-name>
   device: cpu | cuda
   profile: FULL | DECENT | QUICK | DEBUG | VALIDATION
   country-name: for Microsoft, optional, overrides <iso-code>.
   iso-code: Country Alpha2 ISO code: https://en.wikipedia.org/wiki/List_of_ISO_3166_country_codes
      Canada: CA
      US: US
   source: microsoft | osm | google
   calls: https://github.com/microsoft/building-damage-assessment/blob/main/download_building_footprints.py
   buffer: in meters.

🌐STAC Catalog: Maxar Open Data "Satellite imagery for select sudden onset major crisis events"	🏛️Vision Algo: Semantic Segmentation segmentation_models.pytorch
🧑🏽‍🚒Building Damage Analysis using Microsoft, OSM, and Google footprints through microsoft/building-damage-assessment	🧑🏽‍🚒Analytics Damage information for multi-datacube areas.

---

This workflow is inspired by microsoft/building-damage-assessment and palisades buildings download_footprints calls download_building_footprints.py from the same repo - through satellite-image-deep-learning.

---

built by 🌀 blue_options-4.197.1, based on 🧑🏽‍🚒 palisades-4.135.1.