sfn

This subpackage provides the requisite classes and functions for performing surface feature navigation in GIANT.

Description

Surface Feature Navigation (SFN) is a relative OpNav technique where we extract bearing measurements of individual surface features from an image. Typically, we do this for many features in a single image, which in turn allows us to essentially triangulate the camera from a single image. As such it is a very powerful technique for navigation and is generally preferred once targets grow large enough in the navigation images that we can reliably recognize the features on the surface.

There are a number of different ways to do surface feature navigation, but perhaps one of the oldest (from a space use perspective) involves defining features as small patches of surface instead of actual features (like craters/rocks/etc). For each of these patches of surface, we can then predict what they should look like in an image and use cross correlation to locate the actual location in the image. This technique was popularized by the Stereophotoclinometry software suite developed by Dr. Robert Gaskell and conceptually defines how we do SFN in GIANT.

From this package you can import everything you need for doing SFN in GIANT. The most important things that you will need from here are the SurfaceFeatureNavigation class which does the actual navigation, the FeatureCatalogue class which is used to interface a list of features we want to process with the Scene in GIANT, and the VisibleFeatureFinder which is used to determine which features in the catalogue should actually be visible in the images.

For more detailed description (and hints for tuning for SFN) refer to the following module and class documentation for the objects defined in this package.

Use

In general you won’t interact too much with the classes/functions defined in this package, with the exception of the VisibleFeatureFinder, which you will need to use to specify the settings you want when identifying possibly visible features. Beyond that, most of the other tools are automatically handled for you in the RelativeOpNav class. If you need to build a feature catalogue for use with SFN, you should take a look at the spc_to_feature_catalogue and tile_shape scripts which may be able to do it for you.

Modules

sfn_class	This module provides the capability to locate surface features from a target in an image using 2D cross-correlation.
sfn_correlators	This module defines the spatial correlator used by SurfaceFeatureNavigation for subpixel locations of features in an image.
surface_features	This module implements the idea of a surface feature and surface feature catalogue for GIANT.