Superpixel-based depth image super-resolution

Yongseok Soh, Jae Young Sim, Chang-Su Kim, Sang Uk Lee

Research output: Chapter in Book/Report/Conference proceedingConference contribution

12 Citations (Scopus)

Abstract

Due to the development of depth sensors, such as time-of-flight (ToF) cameras, it becomes easier to acquire depth information directly from a scene. Although such devices enable us to obtain depth maps at video frame rates, the depth maps often have low resolutions only. A typical ToF camera retrieves depth maps of resolution 320 × 200, which is much lower than the resolutions of high definition color images. In this work, we propose a depth image super-resolution algorithm, which operates robustly even when there is a large resolution gap between a depth image and a reference color image. To prevent edge smoothing artifacts, which are the main drawback of conventional techniques, we adopt a superpixel-based approach and develop an edge enhancing scheme. Simulation results demonstrate that the proposed algorithm aligns the edges of a depth map to accurately coincide with those of a high resolution color image.

Original languageEnglish
Title of host publicationProceedings of SPIE - The International Society for Optical Engineering
Volume8290
DOIs
Publication statusPublished - 2012 Jun 12
Event3-Dimensional Image Processing (3DIP) and Applications II - Burlingame, CA, United States
Duration: 2012 Jan 242012 Jan 26

Other

Other3-Dimensional Image Processing (3DIP) and Applications II
CountryUnited States
CityBurlingame, CA
Period12/1/2412/1/26

Fingerprint

Depth Map
Super-resolution
Color Image
Time-of-flight
Color
Camera
Cameras
color
Smoothing
High Resolution
cameras
Sensor
Sensors
smoothing
Demonstrate
artifacts
Simulation
high resolution
sensors

Keywords

  • Depth image
  • Depth map super-resolution
  • Image interpolation
  • Superpixel

ASJC Scopus subject areas

  • Applied Mathematics
  • Computer Science Applications
  • Electrical and Electronic Engineering
  • Electronic, Optical and Magnetic Materials
  • Condensed Matter Physics

Cite this

Soh, Y., Sim, J. Y., Kim, C-S., & Lee, S. U. (2012). Superpixel-based depth image super-resolution. In Proceedings of SPIE - The International Society for Optical Engineering (Vol. 8290). [82900D] https://doi.org/10.1117/12.909848

Superpixel-based depth image super-resolution. / Soh, Yongseok; Sim, Jae Young; Kim, Chang-Su; Lee, Sang Uk.

Proceedings of SPIE - The International Society for Optical Engineering. Vol. 8290 2012. 82900D.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Soh, Y, Sim, JY, Kim, C-S & Lee, SU 2012, Superpixel-based depth image super-resolution. in Proceedings of SPIE - The International Society for Optical Engineering. vol. 8290, 82900D, 3-Dimensional Image Processing (3DIP) and Applications II, Burlingame, CA, United States, 12/1/24. https://doi.org/10.1117/12.909848
Soh Y, Sim JY, Kim C-S, Lee SU. Superpixel-based depth image super-resolution. In Proceedings of SPIE - The International Society for Optical Engineering. Vol. 8290. 2012. 82900D https://doi.org/10.1117/12.909848
Soh, Yongseok ; Sim, Jae Young ; Kim, Chang-Su ; Lee, Sang Uk. / Superpixel-based depth image super-resolution. Proceedings of SPIE - The International Society for Optical Engineering. Vol. 8290 2012.
@inproceedings{a760ec79f6614f66a62ae3e749c41374,
title = "Superpixel-based depth image super-resolution",
abstract = "Due to the development of depth sensors, such as time-of-flight (ToF) cameras, it becomes easier to acquire depth information directly from a scene. Although such devices enable us to obtain depth maps at video frame rates, the depth maps often have low resolutions only. A typical ToF camera retrieves depth maps of resolution 320 × 200, which is much lower than the resolutions of high definition color images. In this work, we propose a depth image super-resolution algorithm, which operates robustly even when there is a large resolution gap between a depth image and a reference color image. To prevent edge smoothing artifacts, which are the main drawback of conventional techniques, we adopt a superpixel-based approach and develop an edge enhancing scheme. Simulation results demonstrate that the proposed algorithm aligns the edges of a depth map to accurately coincide with those of a high resolution color image.",
keywords = "Depth image, Depth map super-resolution, Image interpolation, Superpixel",
author = "Yongseok Soh and Sim, {Jae Young} and Chang-Su Kim and Lee, {Sang Uk}",
year = "2012",
month = "6",
day = "12",
doi = "10.1117/12.909848",
language = "English",
isbn = "9780819489371",
volume = "8290",
booktitle = "Proceedings of SPIE - The International Society for Optical Engineering",

}

TY - GEN

T1 - Superpixel-based depth image super-resolution

AU - Soh, Yongseok

AU - Sim, Jae Young

AU - Kim, Chang-Su

AU - Lee, Sang Uk

PY - 2012/6/12

Y1 - 2012/6/12

N2 - Due to the development of depth sensors, such as time-of-flight (ToF) cameras, it becomes easier to acquire depth information directly from a scene. Although such devices enable us to obtain depth maps at video frame rates, the depth maps often have low resolutions only. A typical ToF camera retrieves depth maps of resolution 320 × 200, which is much lower than the resolutions of high definition color images. In this work, we propose a depth image super-resolution algorithm, which operates robustly even when there is a large resolution gap between a depth image and a reference color image. To prevent edge smoothing artifacts, which are the main drawback of conventional techniques, we adopt a superpixel-based approach and develop an edge enhancing scheme. Simulation results demonstrate that the proposed algorithm aligns the edges of a depth map to accurately coincide with those of a high resolution color image.

AB - Due to the development of depth sensors, such as time-of-flight (ToF) cameras, it becomes easier to acquire depth information directly from a scene. Although such devices enable us to obtain depth maps at video frame rates, the depth maps often have low resolutions only. A typical ToF camera retrieves depth maps of resolution 320 × 200, which is much lower than the resolutions of high definition color images. In this work, we propose a depth image super-resolution algorithm, which operates robustly even when there is a large resolution gap between a depth image and a reference color image. To prevent edge smoothing artifacts, which are the main drawback of conventional techniques, we adopt a superpixel-based approach and develop an edge enhancing scheme. Simulation results demonstrate that the proposed algorithm aligns the edges of a depth map to accurately coincide with those of a high resolution color image.

KW - Depth image

KW - Depth map super-resolution

KW - Image interpolation

KW - Superpixel

UR - http://www.scopus.com/inward/record.url?scp=84861925561&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84861925561&partnerID=8YFLogxK

U2 - 10.1117/12.909848

DO - 10.1117/12.909848

M3 - Conference contribution

AN - SCOPUS:84861925561

SN - 9780819489371

VL - 8290

BT - Proceedings of SPIE - The International Society for Optical Engineering

ER -