GIPPY: A high performance geospatial image processing library for Python


Combining the power of GDAL with a C++ image processing library called CImg, GIPPY provides a Python interface to easily perform common image processing tasks on large remote sensing data sets. By automatically and transparently chunking input imagery and allowing chaining together of operations, users can put together algorithms that are then processed efficiently by deferring all operations until read time, while utilizing multi-threaded functionality of the Boost C++ libraries.

GIPPY not only provides a simply Pythonic interface while abstracting away the chunking and processing of the data, it also has a library of common algorithms that can be performed with a minimal of coding. Included are algorithms for mosaicking together multiple images in any combination of projections, calculating spectral correlation and covariance, cloud detection algorithms ACCA and F-mask, multiple indices (NDVI, EVI, LSWI, NDSI, NDWI, SATVI, MSAVI-2), K-means unsupervised classifier, arbitrary linear transforms, RX Detector, and Spectral Matched Filters.

GIPPY is an open-source project available on GitHub and released under the Apache License.

Session details
Speaker(s): Session Type: Experience level:
Track: Tags:
Schedule info
Session Time Slot(s):
Tuesday, May 3, 2016 - 17:40 to 17:55