A Python library that provides efficient batch convex hull computation for 3D points using the qhull library and thread pool parallelization.
- Batch processing of multiple point clouds
- Thread pool for parallel computation
- Built on the robust qhull library
- Install dependencies:
# Install qhull development package
sudo apt-get install libqhull-dev # Ubuntu/Debian
# or
sudo dnf install qhull-devel # Fedora
# or
micromamba install pybind11 libqhull
# Install Python dependencies
pip install -r requirements.txt- Build and install:
pip install .python test_pyqhull.py
Threadpool | Batch Size | Avg Time (s) | Stderr Time (s)
-------------------------------------------------------
1 | 1 | 0.000239 | 0.000007
1 | 2 | 0.000295 | 0.000019
1 | 4 | 0.000342 | 0.000011
1 | 8 | 0.000517 | 0.000017
1 | 16 | 0.000845 | 0.000008
1 | 32 | 0.001392 | 0.000010
1 | 64 | 0.002420 | 0.000024
1 | 128 | 0.004446 | 0.000021
1 | 256 | 0.008318 | 0.000015
1 | 512 | 0.016215 | 0.000038
1 | 1024 | 0.031808 | 0.000045
1 | 2048 | 0.063153 | 0.000054
1 | 4096 | 0.126001 | 0.000113
4 | 1 | 0.000199 | 0.000015
4 | 2 | 0.000132 | 0.000022
4 | 4 | 0.000195 | 0.000017
4 | 8 | 0.000201 | 0.000017
4 | 16 | 0.000374 | 0.000031
4 | 32 | 0.000479 | 0.000028
4 | 64 | 0.000851 | 0.000022
4 | 128 | 0.001408 | 0.000011
4 | 256 | 0.002431 | 0.000016
4 | 512 | 0.004410 | 0.000016
4 | 1024 | 0.008379 | 0.000027
4 | 2048 | 0.016244 | 0.000019
4 | 4096 | 0.032122 | 0.000052
16 | 1 | 0.000136 | 0.000024
16 | 2 | 0.000108 | 0.000005
16 | 4 | 0.000148 | 0.000017
16 | 8 | 0.000159 | 0.000017
16 | 16 | 0.000201 | 0.000021
16 | 32 | 0.000207 | 0.000023
16 | 64 | 0.000400 | 0.000060
16 | 128 | 0.000554 | 0.000023
16 | 256 | 0.000888 | 0.000024
16 | 512 | 0.001486 | 0.000022
16 | 1024 | 0.002599 | 0.000033
16 | 2048 | 0.004648 | 0.000043
16 | 4096 | 0.009094 | 0.000105
64 | 1 | 0.000143 | 0.000012
64 | 2 | 0.000172 | 0.000033
64 | 4 | 0.000146 | 0.000014
64 | 8 | 0.000159 | 0.000018
64 | 16 | 0.000206 | 0.000030
64 | 32 | 0.000590 | 0.000267
64 | 64 | 0.000557 | 0.000105
64 | 128 | 0.000887 | 0.000075
64 | 256 | 0.001027 | 0.000099
64 | 512 | 0.001941 | 0.000317
64 | 1024 | 0.003286 | 0.000277
64 | 2048 | 0.006420 | 0.000389
64 | 4096 | 0.012534 | 0.000199
Threadpool | Batch Size | Avg Time (s) | Stderr Time (s)
-------------------------------------------------------
1 | 1 | 0.000450 | 0.000025
1 | 2 | 0.000632 | 0.000029
1 | 4 | 0.001043 | 0.000014
1 | 8 | 0.001834 | 0.000028
1 | 16 | 0.003349 | 0.000013
1 | 32 | 0.006338 | 0.000036
1 | 64 | 0.012222 | 0.000050
1 | 128 | 0.024085 | 0.000046
1 | 256 | 0.047796 | 0.000080
1 | 512 | 0.095023 | 0.000122
1 | 1024 | 0.189710 | 0.000194
1 | 2048 | 0.378532 | 0.000445
1 | 4096 | 0.756263 | 0.000627
4 | 1 | 0.000400 | 0.000018
4 | 2 | 0.000395 | 0.000032
4 | 4 | 0.000445 | 0.000008
4 | 8 | 0.000657 | 0.000007
4 | 16 | 0.001062 | 0.000017
4 | 32 | 0.001854 | 0.000020
4 | 64 | 0.003397 | 0.000013
4 | 128 | 0.006446 | 0.000023
4 | 256 | 0.012485 | 0.000031
4 | 512 | 0.024312 | 0.000034
4 | 1024 | 0.048147 | 0.000076
4 | 2048 | 0.095628 | 0.000064
4 | 4096 | 0.190567 | 0.000175
16 | 1 | 0.000313 | 0.000024
16 | 2 | 0.000301 | 0.000009
16 | 4 | 0.000325 | 0.000015
16 | 8 | 0.000362 | 0.000022
16 | 16 | 0.000633 | 0.000135
16 | 32 | 0.000701 | 0.000010
16 | 64 | 0.001146 | 0.000025
16 | 128 | 0.001946 | 0.000016
16 | 256 | 0.003505 | 0.000017
16 | 512 | 0.006559 | 0.000025
16 | 1024 | 0.012595 | 0.000018
16 | 2048 | 0.024629 | 0.000025
16 | 4096 | 0.048726 | 0.000051
48 | 1 | 0.000402 | 0.000094
48 | 2 | 0.000316 | 0.000006
48 | 4 | 0.000376 | 0.000015
48 | 8 | 0.000382 | 0.000018
48 | 16 | 0.000472 | 0.000030
48 | 32 | 0.000653 | 0.000094
48 | 64 | 0.001007 | 0.000132
48 | 128 | 0.001138 | 0.000055
48 | 256 | 0.001836 | 0.000074
48 | 512 | 0.003054 | 0.000030
48 | 1024 | 0.005452 | 0.000030
48 | 2048 | 0.010419 | 0.000015
48 | 4096 | 0.020342 | 0.000131
96 | 1 | 0.000339 | 0.000013
96 | 2 | 0.000330 | 0.000005
96 | 4 | 0.000368 | 0.000015
96 | 8 | 0.000444 | 0.000032
96 | 16 | 0.000517 | 0.000030
96 | 32 | 0.000602 | 0.000016
96 | 64 | 0.001198 | 0.000276
96 | 128 | 0.001560 | 0.000310
96 | 256 | 0.001885 | 0.000086
96 | 512 | 0.002863 | 0.000114
96 | 1024 | 0.005077 | 0.000280
96 | 2048 | 0.009011 | 0.000284
96 | 4096 | 0.016376 | 0.000235