High level API Reference

This page provides an overview of all public objects, functions and methods exposed in the high level, public API of orthoseg.

Core entry points

The primary way to use orthoseg from Python is to import and call the main functions from the top-level orthoseg module:

train	Run a training session for the config specified.
predict	Run a prediction for the config specified.
postprocess	Postprocess the output of a prediction for the config specified.
load_images	Load and cache images for a segmentation project.
validate	Run a validating session for the config specified.

The entries above generate the detailed API pages automatically during the Sphinx build. Each function gets its own page under the API toctree, which avoids having multiple indexed descriptions of the same object.

Low-level API note

It is also possible to call low-level orthoseg functions in the orthoseg.lib subpackage directly from Python. However, these are not considered as a public API and may change without notice. Hence, detailed API documentation is not provided for these low-level modules.