Demo project for Semantic3D (semantic-8) segmentation with Open3D and PointNet++. The purpose of this project is to showcase the usage of Open3D in deep learning pipelines and provide a clean baseline implementation for semantic segmentation on Semantic3D dataset. Here's our entry on the semantic-8 test benchmark page.
Open3D is an open-source library that supports rapid development of software that deals with 3D data. The Open3D frontend exposes a set of carefully selected data structures and algorithms in both C++ and Python. The backend is highly optimized and is set up for parallelization. We welcome contributions from the open-source community.
In this project, Open3D was used for
- Point cloud data loading, writing, and visualization. Open3D provides efficient implementations of various point cloud manipulation methods.
- Data pre-processing, in particular, voxel-based down-sampling.
- Point cloud interpolation, in particular, fast nearest neighbor search for label interpolation.
- And more.
This project is forked from Mathieu Orhan and Guillaume Dekeyser's repo, which, is forked from the original PointNet2. We thank the original authors for sharing their methods.
Download the dataset Semantic3D and extract it by running the following commands:
cd dataset/semantic_raw
.
bash download_semantic3d.sh
.
Open3D-PointNet2-Semantic3D/dataset/semantic_raw
├── bildstein_station1_xyz_intensity_rgb.labels
├── bildstein_station1_xyz_intensity_rgb.txt
├── bildstein_station3_xyz_intensity_rgb.labels
├── bildstein_station3_xyz_intensity_rgb.txt
├── ...
Run
python preprocess.py
Open3D is able to read .pcd
files much more efficiently.
Open3D-PointNet2-Semantic3D/dataset/semantic_raw
├── bildstein_station1_xyz_intensity_rgb.labels
├── bildstein_station1_xyz_intensity_rgb.pcd (new)
├── bildstein_station1_xyz_intensity_rgb.txt
├── bildstein_station3_xyz_intensity_rgb.labels
├── bildstein_station3_xyz_intensity_rgb.pcd (new)
├── bildstein_station3_xyz_intensity_rgb.txt
├── ...
Run
python downsample.py
The downsampled dataset will be written to dataset/semantic_downsampled
. Points with
label 0 (unlabled) are excluded during downsampling.
Open3D-PointNet2-Semantic3D/dataset/semantic_downsampled
├── bildstein_station1_xyz_intensity_rgb.labels
├── bildstein_station1_xyz_intensity_rgb.pcd
├── bildstein_station3_xyz_intensity_rgb.labels
├── bildstein_station3_xyz_intensity_rgb.pcd
├── ...
We need to build TF kernels in tf_ops
. First, activate the virtualenv and make
sure TF can be found with current python. The following line shall run without
error.
python -c "import tensorflow as tf"
Then build TF ops. You'll need CUDA and CMake 3.8+.
cd tf_ops
mkdir build
cd build
cmake ..
make
After compilation the following .so
files shall be in the build
directory.
Open3D-PointNet2-Semantic3D/tf_ops/build
├── libtf_grouping.so
├── libtf_interpolate.so
├── libtf_sampling.so
├── ...
Verify that that the TF kernels are working by running
cd .. # Now we're at Open3D-PointNet2-Semantic3D/tf_ops
python test_tf_ops.py
Run
python train.py
By default, the training set will be used for training and the validation set
will be used for validation. To train with both training and validation set,
use the --train_set=train_full
flag. Checkpoints will be output to
log/semantic
.
Pick a checkpoint and run the predict.py
script. The prediction dataset is
configured by --set
. Since PointNet2 only takes a few thousand points per
forward pass, we need to sample from the prediction dataset multiple times to
get a good coverage of the points. Each sample contains the few thousand points
required by PointNet2. To specify the number of such samples per scene, use the
--num_samples
flag.
python predict.py --ckpt log/semantic/best_model_epoch_040.ckpt \
--set=validation \
--num_samples=500
The prediction results will be written to result/sparse
.
Open3D-PointNet2-Semantic3D/result/sparse
├── sg27_station4_intensity_rgb.labels
├── sg27_station4_intensity_rgb.pcd
├── sg27_station5_intensity_rgb.labels
├── sg27_station5_intensity_rgb.pcd
├── ...
The last step is to interpolate the sparse prediction to the full point cloud. We use Open3D's K-NN hybrid search with specified radius.
python interpolate.py
The prediction results will be written to result/dense
.
Open3D-PointNet2-Semantic3D/result/dense
├── sg27_station4_intensity_rgb.labels
├── sg27_station5_intensity_rgb.labels
├── ...
Finally, if you're submitting to Semantic3D benchmark, we've included a handy tools to rename the submission file names.
python renamer.py
dataset/semantic_raw
: Raw Semantic3D data, .txt and .labels files. Also contains the .pcd file generated bypreprocess.py
.dataset/semantic_downsampled
: Generated fromdownsample.py
. Downsampled data, contains .pcd and .labels files.result/sparse
: Generated frompredict.py
. Sparse predictions, contains .pcd and .labels files.result/dense
: Dense predictions, contains .labels files.result/dense_label_colorized
: Dense predictions with points colored by label type.