Line and Plane based Indoor RGB-D Compass (LPIC)

This package provides a MATLAB implementation of CVPR 2018 paper: "Indoor RGB-D Compass from a Single Line and Plane" for the purpose of research and study only. Note that this repository only includes simplified proposed 3-DoF rotation tracking example codes to understand how the LPIC works in structured environments.

LPIC

1. Goal

Our goal is to estimate 3-DoF camera orientation with respect to the spatial regularities of indoor structured environments. The proposed LPIC estimates absolute 3-DoF camera orientation from only a single line and a single plane, which corresponds to the theoretical minimal sampling for 3-DoF rotation estimation. Our algorithm requires a plane and a line on the plane aligned with the Manhattan world (MW) to be visible, which is typically the case in most indoor environments.

LPIC

2. Prerequisites

This package is tested on the MATLAB R2019b on Windows 7 64-bit. Some of the functions such as estimateSurfaceNormalGradient_mex.mexw64 are compiled as MEX file to speed up the computation. You can use estimateSurfaceNormalGradient.m instead if you cannot compile MEX file in your OS.

3. Usage

Download the TUM-RGBD dataset from https://vision.in.tum.de/data/datasets/rgbd-dataset.
Or, Use the TUMRGBDdataset/rgbd_dataset_freiburg3_structure_notexture_far/ included in this package.
Define 'datasetPath' correctly in your directory at setupParams_TUM_RGBD.m file.
Run LPIC_core/main_script_TUM_RGBD.m, which will give you the 3-DoF camera orientation tracking result. Enjoy! :)

4. Publications

The approach is described and used in the following publications:

Indoor RGB-D Compass from a Single Line and Plane (Pyojin Kim, Brian Coltin, and H. Jin Kim), CVPR 2018.

You can find more related papers at http://pyojinkim.com/_pages/pub/index.html.

5. License

The package is licensed under the MIT License, see http://opensource.org/licenses/MIT.

if you use LPIC in an academic work, please cite:

@inproceedings{kim2018indoor,
  author = {Kim, Pyojin and Coltin, Brian and Kim, H Jin},
  title = {Indoor RGB-D Compass from a Single Line and Plane},
  year = {2018},
  booktitle = {IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
 }

PyojinKim / LPIC

readme