estimate-depth

The estimate-depth tool estimates a series of depth maps for video frames based off of known or calculated cameras associated with those frames.

The default configuration of this tool depends on algorithm implementations in the Super3d and VTK arrows which will only available if the KWIVER_ENABLE_SUPER3D and KWIVER_ENABLE_VTK CMake flags are enabled.

kwiver estimate-depth       [options]

Required arguments:

video-source - name of input video file.

input-cameras-dir - name of the directory containing the krtd camera files(default: results/krtd)

output-depths-dir - name of the directory to write depth maps to (default: results/depth)

Options are:

-h, --help

Display applet usage information.

-c, --config arg

Configuration file for tool

-o, --output-config arg

Output a configuration. This may be seeded with a configuration file from -c/–config.

-f, --frame arg

The frame number to compute depth for.

-l, --input-landmarks-file arg

3D sparse features

-m, --mask-source arg

Masks of unusable regions

Default configuration

# Default configuration for the "compute_depth" algorithm used in the applet

block compute_depth
  include super3d_depth_map.conf
  # number of support images for each depth map
  num_support = 20
  # select support images from the set of cameras with view angles
  # within this many degrees of the reference image
  angle_span = 15
endblock

block batch_depth
	first_frame = 0
	end_frame = -1
	num_depth = 20
endblock

# Customization of the default parameters for the applet
# ---------------------------------------------------