Overview

A suite of functionalities for mass-processing of satellite imagery in Google Earth Engine. It has been created to facilitate pre-processing and time series analyses of Landsat, Sentinel-2 and NICFI PlanetScope data. Key functionalities vary across sensors but include

• cross-sensor collection integration (Landsat C1)
• cloud-masking (Landsat C1, Sentinel-2)
• co-registration using image.displace() (PlanetScope ~ Sentinel-2)
• spectral-temporal metrics (Landsat, Sentinel-2, PlanetScope)
• glcm texture metrics (Landsat, Sentinel-2, PlanetScope)
• facilitated image-wise exports of ee.ImageCollections to Google Drive
• conversion of local vector files to ee.FeatureGeometry or ee.FeatureCollections

Structure

• src/ eepypr core functions
• app/ application examples
  • nicfi stm generate spectral-temporal metrics from NICFI PlanetScope mosaics with optional coregistration based on Sentinel-2
  • aralsea_stm spectral-temporal metrics from Landsat Collection 1 data
  • crop_types crop type & intercropping mapping in Oyo state, Nigeria

References

The functions in this module were created and used in research conducted at Humboldt-Universität zu Berlin (Germany) and UCLouvain (Belgium), such as a paper on large-area mapping of smallholder agriculture, or mapping post-Soviet cropping practices upstream of the Aral Sea basin. For further information read the function descriptions or contact philippe.rufin@uclouvain.be.

Suggested Alternatives

I warmly recommend using the richer and more user-friendly GEEO library developed by my colleague Leon Nill: https://github.com/leonsnill/geeo/