Depth Image
Enhancement Using Local Tangent Plane Approximations
Source measured depth image |
Resulting enhanced depth image |
Abstract:
This paper describes a
depth image enhancement method for consumer RGB-D
cameras. Most existing methods use the pixel-coordinates of the aligned
color image. Because the image plane generally has no relationship to
the measured surfaces, the global coordinate system is not suitable to
handle their local geometries. To improve enhancement accuracy, we use
local tangent planes as local coordinates for the measured surfaces.
Our method is composed of two steps, a calculation of the local
tangents and surface reconstruction. To accurately estimate the local
tangents, we propose a color heuristic calculation and an orientation
correction using their positional relationships. Additionally, we
propose a surface reconstruction method by ray-tracing to local
tangents. In our method, accurate depth image enhancement is achieved
by using the local geometries approximated by the local tangents.
Videos:
Algorithm overview :
[ high
resolution (MP4, 47.8 MB),
low
resolution (MP4, 5.4 MB) ]
Evaluation on real sensor data : [ high resolution (MP4, 42.0 MB), low resolution (MP4, 5.0 MB) ]
Evaluation on real sensor data : [ high resolution (MP4, 42.0 MB), low resolution (MP4, 5.0 MB) ]
Publication:
Depth Image Enhancement Using Local
Tangent Plane Approximations, Kiyoshi Matsuo and Yoshimitsu
Aoki.
In Computer Vision and Pattern Recognition (CVPR), 2015. [ Preprint, Extended abstract, Supplementary materials, Poster ]
In Computer Vision and Pattern Recognition (CVPR), 2015. [ Preprint, Extended abstract, Supplementary materials, Poster ]
Related talk:
Depth Image Enhancement Using Local
Tangent Plane Approximations, Kiyoshi Matsuo and Yoshimitsu
Aoki.
( Invited talk in the 18th Meeting on Image Recognition and Understanding (MIRU). July 29, 2015. ) [ Slides, Poster ]
( Invited talk in the 18th Meeting on Image Recognition and Understanding (MIRU). July 29, 2015. ) [ Slides, Poster ]
Depth Image Enhancement Using Local Tangent Plane Approximations, Kiyoshi Matsuo and Yoshimitsu Aoki.
( Talk in the 17th Image Processing Tokyo. November 26, 2015. ) [ Slides ]
Source code:
Original
codes used in the paper can not be distributed by some regulations and
I could not share any codes of this algorithm. ( I'm sorry... )
I could not share any codes of this algorithm. ( I'm sorry... )
Bibtex:
@InProceedings{Matsuo_2015_CVPR,
author = {Matsuo, Kiyoshi and Aoki, Yoshimitsu},
title = {Depth Image Enhancement Using Local Tangent Plane Approximations},
booktitle = {The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},
year = {2015}
}
Contact:
Kiyoshi Matsuo (primary contact) | [URL] | [e-mail] |
Yoshimitsu Aoki | [URL] | [e-mail] |
Copyright
(C) 2015 Kiyoshi
Matsuo.
E-mail: ki4ma20[at]gmail.com
Hokuyo Automatic
Co., LTD | Aoki
Media Sensing Lab.