Clutterbox Experiment Implementation Source Code

This source code repository accompanies the paper 'Radial Intersection Count Image: a Clutter Resistant 3D Shape Descriptor', and contains the implementation of the proposed clutterbox experiment described in it.

Compiling

The project has been developed on Ubuntu Linux 18.04 LTS. It should in principle compile for Windows and MacOS installations too, although I have not personally tested this.

For compiling on Ubuntu Linux, the following Aptitude packages were used (package versions used for results in paper are listed):

• cmake (version 3.10.2)
• libeigen3-dev (version 3.3.4-4)
• libpcl-dev (version 1.8.1)
• python3 (version 3.6.7-1~18.04)
• nvidia-cuda-toolkit (version 10.1.105)
• nvidia-cuda-dev (version 10.1.105)
• gcc-7 (version 7.4.0-1)

Compiling can be done by running the following from the project's root directory:

mkdir build
cd build
cmake .. -DCMAKE_BUILD_TYPE=Release
make -j 4

The project generates several executables:

• clutterbox: Main executable for running the Clutterbox experiment
• projectionBenchmark: The paper discusses a microbenchmark for comparing the PCL method for projecting points into cylindrical coordinates relative to our proposed method. This executable performs this benchmark.
• clutterEstimator: Tool used to generate the heatmaps comparing search result ranks to clutter levels in the area around a given point. This tool was used to estimate the clutter level around each point in a scene.
• libShapeDescriptor/imagerenderer: Tool for using our QUICCI, RICI, and SI implementations to render images of a given input object. Outputs PNG images.

Credits

• Development and implementation: Bart Iver van Blokland, NTNU Visual Computing Lab
• Supervision: Theoharis Theoharis, NTNU Visual Computing Lab

If you use (parts of) this library in your research, we kindly ask you reference the papers on which this project is based:

@article{van2020radial,
    title={Radial intersection count image: A clutter resistant 3D shape descriptor},
    author={van Blokland, Bart Iver and Theoharis, Theoharis},
    journal={Computers \& Graphics},
    volume="91",
    pages="118--128",
    year={2020},
    publisher={Elsevier}
}

@article{van2020indexing,
    title={An Indexing Scheme and Descriptor for 3D Object Retrieval Based on Local Shape Querying},
    author={van Blokland, Bart Iver and Theoharis, Theoharis},
    journal={Computers \& Graphics},
    volume="92",
    pages="55-66",
    year={2020},
    publisher={Elsevier}
}