When I try to run the 3DMASC -> train classifier through CloudCompare (ver. 2.13.2), I fail when trying to include the context based features DZx and DHy.
I have read the manual, but still cannot figure out how to correctly structure the config file.
This is how it looks like at the moment,
# NAMES OF YOUR POINT CLOUDS
cloud: PC1=railway_000033_rail.laz # Point cloud for training
cloud: corepoints=railway_000033_rail_sub04.laz
# CORE POINTS
core_points: corepoints
# Feature points
cloud: CTX=railway_000033_rail.laz
# SCALES
scales: 0.5;1;1.5;2 # scales as a list
# FEATURES
feature: DIP_SCx_PC1
feature: DIPDIR_SCx_PC1
feature: PCA1_SCx_PC1
feature: PCA2_SCx_PC1
feature: PCA3_SCx_PC1
feature: SPHER_SCx_PC1
feature: ANISO_SCx_PC1
feature: LINEA_SCx_PC1
feature: PLANA_SCx_PC1
feature: Zmin_SCx_PC1
feature: Zmax_SCx_PC1
feature: ZRANGE_SCx_PC1
feature: INT_SCx_MEDIAN_PC1
feature: DZx_SCx_MEDIAN_PC1
#feature: DZx_SCx_MEDIAN_CTX
#feature: DHy_SCx_MEDIAN_CTX
#We already calculated the verticality, you can ignore it if an error shows
#feature: VERT_SCx_MEDIAN_PC1
When I run that I get,
Malformed file: expecting a valid and positive number after 'DZ' on line #36
Unable to make use of context based features
Re: Unable to make use of context based features
Ok, I figured it out.
This is how you structure the config file,
# NAMES OF YOUR POINT CLOUDS
cloud: PC1=railway_000033_rail.laz # Point cloud for training
cloud: corepoints=railway_000033_rail_sub04.laz
# CORE POINTS
core_points: corepoints
# Feature points
cloud: CTX=railway_000033_rail.laz
# SCALES
scales: 0.5;1;1.5;2 # scales as a list
# FEATURES
feature: DIP_SCx_PC1
feature: DIPDIR_SCx_PC1
feature: PCA1_SCx_PC1
feature: PCA2_SCx_PC1
feature: PCA3_SCx_PC1
feature: SPHER_SCx_PC1
feature: ANISO_SCx_PC1
feature: LINEA_SCx_PC1
feature: PLANA_SCx_PC1
feature: Zmin_SCx_PC1
feature: Zmax_SCx_PC1
feature: ZRANGE_SCx_PC1
feature: INT_SCx_MEDIAN_PC1
feature: DZ10_SC0_corepoints_CTX_2
feature: DZ1_SC0_corepoints_CTX_2
#We already calculated the verticality, you can ignore it if an error shows
#feature: VERT_SCx_MEDIAN_PC1
Here is a more thorough example for anyone else running into this issue,
https://lidar.univ-rennes.fr/sites/lida ... params.txt
This is how you structure the config file,
# NAMES OF YOUR POINT CLOUDS
cloud: PC1=railway_000033_rail.laz # Point cloud for training
cloud: corepoints=railway_000033_rail_sub04.laz
# CORE POINTS
core_points: corepoints
# Feature points
cloud: CTX=railway_000033_rail.laz
# SCALES
scales: 0.5;1;1.5;2 # scales as a list
# FEATURES
feature: DIP_SCx_PC1
feature: DIPDIR_SCx_PC1
feature: PCA1_SCx_PC1
feature: PCA2_SCx_PC1
feature: PCA3_SCx_PC1
feature: SPHER_SCx_PC1
feature: ANISO_SCx_PC1
feature: LINEA_SCx_PC1
feature: PLANA_SCx_PC1
feature: Zmin_SCx_PC1
feature: Zmax_SCx_PC1
feature: ZRANGE_SCx_PC1
feature: INT_SCx_MEDIAN_PC1
feature: DZ10_SC0_corepoints_CTX_2
feature: DZ1_SC0_corepoints_CTX_2
#We already calculated the verticality, you can ignore it if an error shows
#feature: VERT_SCx_MEDIAN_PC1
Here is a more thorough example for anyone else running into this issue,
https://lidar.univ-rennes.fr/sites/lida ... params.txt
-
- Posts: 40
- Joined: Tue Dec 01, 2020 1:21 pm
Re: Unable to make use of context based features
"feature: DZ10_SC0_corepoints_CTX_2" is an old version of the syntax (still valid but deprecated because confusing as features are always computed at the core point).
You should better write "feature: DZ10_SC0_CTX_2".
The reference document for the syntax is here:
https://lidar.univ-rennes.fr/sites/lida ... syntax.pdf
Later evolution will be reported in it if any.
You should better write "feature: DZ10_SC0_CTX_2".
The reference document for the syntax is here:
https://lidar.univ-rennes.fr/sites/lida ... syntax.pdf
Later evolution will be reported in it if any.