模型可视化>> GOOD_descriptor>> 返回
项目作者： SeyedHamidreza

项目描述 ：
Global Orthographic Object Descriptor (GOOD)
高级语言： C++
项目地址: git://github.com/SeyedHamidreza/GOOD_descriptor.git
创建时间： 2017-06-16T13:14:03Z
项目社区：https://github.com/SeyedHamidreza/GOOD_descriptor
开源协议：BSD 3-Clause "New" or "Revised" License
下载


GOOD: A Global Orthographic Object Descriptor for 3D Object Recognition and Manipulation

Hamidreza Kasaei, Ana Maria Tomé, Luís Seabra Lopes, Miguel Oliveira

#

alt text

The Global Orthographic Object Descriptor (GOOD) has been designed to be robust, descriptive and efficient to compute and use. GOOD descriptor has two outstanding characteristics:

  1. Providing a good trade-off among :
    • descriptiveness,
      • robustness,
      • computation time,
      • memory usage.
  2. Allowing concurrent object recognition and pose estimation for manipulation.

The performance of the proposed object descriptor is compared with the main state-of-the-art descriptors. Experimental results show that the overall classification performance obtained with GOOD is comparable to the best performances obtained with the state-of-the-art descriptors. Concerning memory and computation time, GOOD clearly outperforms the other descriptors. Therefore, GOOD is especially suited for real-time applications.
The current implementation of GOOD descriptor supports several functionalities for 3D object recognition and Object Manipulation.

CITING

This is an implementation of the GOOD descriptor, which has been presented in the following papers.
Please adequately refer to the papers any time this code is being used.
If you do publish a paper where GOOD descriptor helped your research, we encourage you to cite the following papers in your publications.

  1. @article{GOODPRL,
  2.     title = "GOOD: A global orthographic object descriptor for 3D object recognition and manipulation",
  3.     journal = "Pattern Recognition Letters",
  4.     volume = "83",
  5.     pages = "312 - 320",
  6.     year = "2016",
  7.     note = "Efficient Shape Representation, Matching, Ranking, and its Applications",
  8.     issn = "0167-8655",
  9.     doi = "http://dx.doi.org/10.1016/j.patrec.2016.07.006",
  10.     url = "http://www.sciencedirect.com/science/article/pii/S0167865516301684",
  11.     author = "S. Hamidreza Kasaei and Ana Maria Tomé and Luís Seabra Lopes and Miguel Oliveira",}
  13. @INPROCEEDINGS{GOODIROS, 
  14.     author={S. H. Kasaei and L. Seabra Lopes and A. M. Tomé and M. Oliveira}, 
  15.     booktitle={2016 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)}, 
  16.     title={An orthographic descriptor for 3D object learning and recognition}, 
  17.     year={2016}, 
  18.     pages={4158-4163}, 
  19.     doi={10.1109/IROS.2016.7759612}, 
  20.     url="http://ieeexplore.ieee.org/document/7759612/"
  21.     month={Oct},}

LICENSE

The GOOD descriptor code is released under the BSD License. A version of this code under a different licensing agreement, intended for commercial use, is also available. Please contact if interested.

DEPENDENCIES

The required dependencies are Point Cloud Library, Boost and Eigen:

  1. PCL (http://www.pointclouds.org/downloads/)
    1.  sudo add-apt-repository ppa:v-launchpad-jochen-sprickerhof-de/pcl
    2.  sudo apt-get update
    3.  sudo apt-get install libpcl-all
  2. Eigen (http://eigen.tuxfamily.org/index.php?title=Main_Page)
    1.  hg clone https://bitbucket.org/eigen/eigen/
    2.  cd build
    3.  cmake ..
    4.  sudo make install

BUILD

We provide a CMakeLists.txt, therefore, the code can be build over different compilers and platforms by making use of CMake.

  1. mkdir build
  2. cd build
  3. cmake ..
  4. make

CODE

In the current distribution you can find:

  • an implementation of the GOOD descriptor (good.h and good.cpp)
  • a sample test code (test_GOOD_descriptor.cpp) that demonstrates how to use it on a test object

How to run the sample code

First of all you will need a point cloud in PCD or PLY format. For your convenience, we have provided a set of sample point clouds in both PLY and PCD format in a folder named “point_cloud”.

The syntax for runing the sample code is: 

  1. ./test_GOOD_descriptor <path/file_name.pcd> [--nogui]    
  2. ./test_GOOD_descriptor <path/file_name.ply> [--nogui]
  • Use [—nogui] option to disable visualization.

Example:

  1. ./test_GOOD_descriptor ../point_cloud/vase.pcd 
  2. ./test_GOOD_descriptor ../point_cloud/vase.pcd --nogui

REAl-TIME DEMONSTRATION

To show all the functionalities and properties of the GOOD descriptor, a demonstration in real time was performed.

:movie_camera: A video of this demonstration is available in: https://youtu.be/iEq9TAaY9u8

CONTACT INFORMATION

  1. Please use the following email addresses if you have questions or want to contribute to this project:
  2. Follow us on our website for updates and to get information about other works:
    • :satellite: www.ai.rug.nl/hkasaei