tile_shape¶
Tile a shape model into SurfaceFeatures in a FeatureCatalogue.
This script tiles an existing shape model in a GIANT format (shapes, kdtree, surface_feature,
etc) into a FeatureCatalgoue of SurfaceFeatures. The features are sampled uniformly from the
existing shape as square grids with a specified size and ground sample distance. Multiple GSDs can be requested in a
single run. The resulting features can also optionally be exported to OBJs or SPC Maplets/Landmarks.
Typically this script will be used in 2 ways. First, if there is a global high resolution of model of the target being
observed (such as from LiDAR data) then this script can be used to tile the high resolution model into features with
varying ground sample distances for direct use in sfn or SPC. Second, this script may be used when building a
shape model using only optical data. After constructing an initial low resolution shape model using limbs, this script
can then be used to tile the low reduction model into initial features that will subsequently be refined through SPC or
other shape modelling techniques.
Warning
For large number of features or when starting with a high-resolution shape model this script can take a long time to run. Therefore we encourage you to use nohup to allow it to complete unhindered.
Warning
This script load/saves some results from/to python pickle files. Pickle files can be used to execute arbitrary code, so you should never open one from an untrusted source.
Usage¶
Generate a feature catalog for Surface Feature Navigation (SFN) containing by tiling a global shape model
usage: tile_shape [-h] [-f FEATURE_OUTPUT] [-c CATALOGUE_OUTPUT] [-m] [-p] [-o OBJS] [-g GSDS [GSDS ...]] [-s SIZE] [-r SHAPE_RADIUS] [-v OVERLAP] shape
Positional Arguments¶
- shape
path to the shape file directory
Named Arguments¶
- -f, --feature_output
The directory to save the feature results to
Default: “./features”
- -c, --catalogue_output
The directory to save the feature results to
Default: “./feature_catalogue.pickle”
- -m, --memory_efficient
use memory efficient triangles
Default: False
- -p, --spc
Make spc stuff
Default: False
- -o, --objs
directory to output objs for each feature to this location (leave none to not make objs)
- -g, --gsds
The gsds to build features at in meters
Default: [1.5, 1, 0.75, 0.35, 0.15]
- -s, --size
The size of each side of a feature
Default: 201
- -r, --shape_radius
The average radius of the shape in meters. Used to determine the number of tiles to make.
Default: 250
- -v, --overlap
The minimum overlap between each tile as a faction in [0, 1). 0 indicates no minimum overlap (default) and 1 would indicate full overlap though that is not possible
Default: 0
WARNING: This script loads/saves some results from/to python pickle files. Pickle files can be used to execute arbitrary code, so you should never open one from an untrusted source.