Skip to content

bartvbl/Quick-Intersection-Count-Change-Image-Reproduction

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

92 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Source code repository for the paper "An Indexing Scheme and Descriptor for 3D Object Retrieval Based on Local Shape Querying"


Notice: This repository is for archival and reproduction purposes only.

Please refer to the following repositories for updated code and documentation:

libShapeDescriptor

Clutterbox

Hamming Tree


This repository contains:

  • A reference implementation of the Quick Intersection Count Change Image (along with the other tested descriptors in the paper's evaluation)
  • A script which can be used to completely reproduce all results presented in the paper

Instructions

You only need to have python 3 installed, which comes with Ubuntu. Any dependencies needed by the project itself can be installed using the menu system in the script itself.

You can run the script by executing:

python3 replicate.py

From the root of the repository.

Refer to the included Manual PDF for further instructions.

System Requirements

The RAM and Disk space requirements are only valid when attempting to reproduce the presented results.

The codebase should be able to compile on Windows, but due to some CUDA driver/SDK compatbility issues we have not yet been able to verify this.

Type Requirements
CPU Does not matter
RAM At least 32GB
Disk Must have about ~120GB of storage available to store the downloaded datasets
GPU Any NVIDIA GPU (project uses CUDA)
OS Ubuntu 16 or higher. Project has been tested on 18 and 20.

Credits

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}
}

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published