Superpixel hierarchy

Xing Wei, Qingxiong Yang, Yihong Gong, Narendra Ahuja, Ming Hsuan Yang

Research output: Contribution to journalArticlepeer-review

36 Citations (Scopus)

Abstract

Superpixel segmentation has been one of the most important tasks in computer vision. In practice, an object can be represented by a number of segments at finer levels with consistent details or included in a surrounding region at coarser levels. Thus, a superpixel segmentation hierarchy is of great importance for applications that require different levels of image details. However, there is no method that can generate all scales of superpixels accurately in real time. In this paper, we propose the superhierarchy algorithm which is able to generate multi-scale superpixels as accurately as the state-of-the-art methods but with one to two orders of magnitude speed-up. The proposed algorithm can be directly integrated with recent efficient edge detectors to significantly outperform the state-of-the-art methods in terms of segmentation accuracy. Quantitative and qualitative evaluations on a number of applications demonstrate that the proposed algorithm is accurate and efficient in generating a hierarchy of superpixels.

Original languageEnglish
Pages (from-to)4838-4849
Number of pages12
JournalIEEE Transactions on Image Processing
Volume27
Issue number10
DOIs
Publication statusPublished - 2018 Oct

Bibliographical note

Funding Information:
This work was supported in part by the State Key Program of National Natural Science Foundation of China under Grant 61332018, in part by the National Basic Research Program of China under Grant 2015CB351705, in part by NSF CAREER under Grant 1149783, in part by Adobe and Nvidia and in part by the Office of Naval Research under Grant N00014-16-1-2314.

Funding Information:
59 10.1109/TIP.2018.2836300 0b00006487d32226 Active orig-research F T F F F F F Publish 10 IEEE 1057-7149 © 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information. Superpixel segmentation has been one of the most important tasks in computer vision. In practice, an object can be represented by a number of segments at finer levels with consistent details or included in a surrounding region at coarser levels. Thus, a superpixel segmentation hierarchy is of great importance for applications that require different levels of image details. However, there is no method that can generate all scales of superpixels accurately in real time. In this paper, we propose the superhierarchy algorithm which is able to generate multi-scale superpixels as accurately as the state-of-the-art methods but with one to two orders of magnitude speed-up. The proposed algorithm can be directly integrated with recent efficient edge detectors to significantly outperform the state-of-the-art methods in terms of segmentation accuracy. Quantitative and qualitative evaluations on a number of applications demonstrate that the proposed algorithm is accurate and efficient in generating a hierarchy of superpixels. Wei, X. Xing Wei Xing Xing Wei Wei Yang, Q. Qingxiong Yang Qingxiong Qingxiong Yang Yang Gong, Y. Yihong Gong Yihong Yihong Gong Gong Ahuja, N. Narendra Ahuja Narendra Narendra Ahuja Ahuja Yang, M. Ming-Hsuan Yang Ming-Hsuan Ming-Hsuan Yang Yang mhyang@ucmerced.edu 2018 Oct. 2018 5 16 2018 6 28 4151601 08360136.pdf 1-1 8360136 Image segmentation Clustering algorithms Vegetation Task analysis Merging Topology Partitioning algorithms Superpixel segmentation Borůvka algorithm National Natural Science Foundation of China 10.13039/501100001809 61332018 National Basic Research Program of China 2015CB351705 National Science Foundation 10.13039/100000001 1149783 Adobe Nvidia 10.13039/100007065 Office of Naval Research 10.13039/100000006 N00014-16-1-2314

Publisher Copyright:
© 2018 IEEE.

All Science Journal Classification (ASJC) codes

  • Software
  • Computer Graphics and Computer-Aided Design

Fingerprint

Dive into the research topics of 'Superpixel hierarchy'. Together they form a unique fingerprint.

Cite this