v0.3.7

Patches and Minor refactor

🚀 new features

• new progress indicator based on alive_progress

argument	hint	default
-h or --help	show help message and exit
-V or --version	show program's version number and exit
-v or --verbose	increase output verbosity
-i or --cbid	force color by id (if color components are parsed)
-c or --cfg [PATH]	path to the config file	auto detect in tree
-f or --frac [F] *	fraction of points for downsampling
-r or --voxel-size [S] *	voxel size for downsampling
-d or --downsample	feed back downsample to the saved point cloud	render only
-s or --save [PATH]	path to .npy file	do not save scene
-p or --make-parent	create parent directories if needed (for --save)
--no-exe	do not execute the app (if --save)
--only [(<=?N)|(N(-N)?)(,\s*N(-N)?)*]	only parse some entries of the config file	parse all entries

python pcv.py --cfg <custom_load>.json5 -vips out/point_cloud.npy -dr10 --only "<=2,4,9-11"

loads the json5 config file (but only with entries n°1,2,4,9,10,11), verbose mode, use id to color points, downsample to have 1 point each 10 units and save result into .npy file and create parent directory if needed

json5 model is as follow (note that only x, y and z are mandatory in "pattern", "pattern" and "skip_first_line" can be overwritten in "configs", and that only "file_path" is not set by default):

{
  "default": {
    "pattern": "{?},{x},{y},{z},{r},{g},{b},{id}",
    "skip_first_line": true
  },
  "configs": [
    {
      "file_path": "<path to file n°1>.csv",
      "source_xyz": [<x1>, <y1>, <z1>]
    },
    {
      "file_path": "<path to file n°2>.csv",
      "source_xyz": [<x2>, <y2>, <z2>]
    },
    ...
  ]
}

⚙️ Builds

curl https://github.com/ThomasByr/point-cloud-visualizer/archive/refs/tags/v0.3.7.zip
unzip point-cloud-visualizer-0.3.7.zip -d point-cloud-visualizer
cd point-cloud-visualizer

conda env create -f environment.yml
conda activate o3d
# or pip install -r requirements.txt

v0.3.5

Next iteration of the first primary version ⚡

all intermediate versions from 0.3.0 to 0.3.5 are fine to use

🚀 new features

• "{X}", "{Y}" and "{Z}" to the pattern for the config file
• --voxel-size as an alternative to --frac
• dowlsampling can be forwarded to the saved .npy file
• cli speedup and better args checking

argument	hint	default
-h or --help	show help message and exit
-V or --version	show program's version number and exit
-v or --verbose	increase output verbosity
-i or --cbid	force color by id (if color components are parsed)
-c or --cfg [PATH]	path to the config file	auto detect in tree
-f or --frac [F] *	fraction of points for downsampling
-r or --voxel-size [S] *	voxel size for downsampling
-d or --downsample	feed back downsample to the saved point cloud	render only
-s or --save [PATH]	path to .npy file	do not save scene
-p or --make-parent	create parent directories if needed (for --save)
--no-exe	do not execute the app (if --save)
--only [(<=?N)|(N(-N)?)(,\s*N(-N)?)*]	only parse some entries of the config file	parse all entries

python pcv.py --cfg <custom_load>.json5 -vips out/point_cloud.npy -dr10 --only "<=2,4,9-11"

loads the json5 config file (but only with entries n°1,2,4,9,10,11), verbose mode, use id to color points, downsample to have 1 point each 10 units and save result into .npy file and create parent directory if needed

json5 model is as follow (note that only x, y and z are mandatory in "pattern", "pattern" and "skip_first_line" can be overwritten in "configs", and that only "file_path" is not set by default):

{
  "default": {
    "pattern": "{?},{x},{y},{z},{r},{g},{b},{id}",
    "skip_first_line": true
  },
  "configs": [
    {
      "file_path": "<path to file n°1>.csv",
      "source_xyz": [<x1>, <y1>, <z1>]
    },
    {
      "file_path": "<path to file n°2>.csv",
      "source_xyz": [<x2>, <y2>, <z2>]
    },
    ...
  ]
}

⚙️ Builds

curl https://github.com/ThomasByr/point-cloud-visualizer/archive/refs/tags/v0.3.5.zip
unzip point-cloud-visualizer-0.3.5.zip -d point-cloud-visualizer
cd point-cloud-visualizer

conda env create -f environment.yml
conda activate o3d
# or pip install -r requirements.txt

v0.3.0

First stable release ⚡

Please do not use anything between 0.1.4 and 0.2.2

🚀 Features

• fast ndimentionnal point representation and loading (3 million points under 12s in txt format!)
• preprocessing step included with numpy formats
• support for python 3.8 to 3.10, minimal and fixed dependencies

argument	hint	default
-h or --help	show help message and exit
-V or --version	show program's version number and exit
-v or --verbose	increase output verbosity
-i or --cbid	force color by id (if color components are parsed)
-c or --cfg [PATH]	path to the config file	auto detect in tree
-f or --frac [F]	fraction of points to render (does not affect saving)	1.0
-s or --save [PATH]	path to .npy file	do not save scene
-p or --make-parent	create parent directories if needed (for --save)
--no-exe	do not execute the app (if --save)
--only [(<=)?N{[,-]N}*]	only parse some entries of the config file	parse all entries

python pcv.py --cfg <custom_load>.json5 -vips out/point_cloud.npy --only "<=2,4,9-11"

loads the json5 config file (but only with entries n°1,2,4,9,10,11), verbose mode, use id to color points, save result into .npy file and create parent directory if needed (pcv.py mode is 744)

json5 model is as follow (note that only x, y and z are mandatory in "pattern", "pattern" and "skip_first_line" can be overwritten in "configs", and that only "file_path" is not set by default):

{
  "default": {
    "pattern": "{?},{x},{y},{z},{r},{g},{b},{id}",
    "skip_first_line": true
  },
  "configs": [
    {
      "file_path": "<path to file n°1>.csv",
      "source_xyz": [<x1>, <y1>, <z1>]
    },
    {
      "file_path": "<path to file n°2>.csv",
      "source_xyz": [<x2>, <y2>, <z2>]
    },
    ...
  ]
}

⚙️ Builds

curl https://github.com/ThomasByr/point-cloud-visualizer/archive/refs/tags/v0.3.0.zip
unzip point-cloud-visualizer-0.3.0.zip -d point-cloud-visualizer
cd point-cloud-visualizer

conda env create -f environment.yml
conda activate o3d
# or pip install -r requirements.txt