Superpixel Matching Based Image Retrieval

Zhixiang He, Xiaoli Sun, Chenhui Li, George Baciu, Yushi Li

Research output: Chapters, Conference Papers, Creative and Literary WorksRGC 32 - Refereed conference paper (with host publication)peer-review

Abstract

Local features of images have been widely used in image retrieval, however, the cost is so heavy. To address this issue, a superpixel-based approach for image retrieval is proposed. We first extract the image structure that preserves the main information and removes the redundant information from the image by smoothing and oversegment a smoothed image into a certain number of superpixels. We then extract the positive candidate superpixels by combining superpixels with local descriptors. Finally, we compute the similarity of two images by analyzing two sets of positive candidate superpixels. Experiments on dataset PQ7 demonstrate the performance of the proposed approach.
Original languageEnglish
Title of host publicationICVIP 2017 : Proceedings of the International Conference on Video and Image Processing
PublisherAssociation for Computing Machinery
Pages156-160
ISBN (Print)9781450353830
DOIs
Publication statusPublished - 27 Dec 2017
Event2017 International Conference on Video and Image Processing (ICVIP 2017) - Nanyang Executive Centre, Singapore, Singapore
Duration: 27 Dec 201729 Dec 2017
http://www.icvip.org/Full_Schedule.pdf

Conference

Conference2017 International Conference on Video and Image Processing (ICVIP 2017)
Abbreviated titleICVIP 2017
PlaceSingapore
CitySingapore
Period27/12/1729/12/17
Internet address

Bibliographical note

Full text of this publication does not contain sufficient affiliation information. With consent from the author(s) concerned, the Research Unit(s) information for this record is based on the existing academic department affiliation of the author(s).

Research Keywords

  • Image retrieval
  • Key-point
  • Smoothing
  • Superpixel

Fingerprint

Dive into the research topics of 'Superpixel Matching Based Image Retrieval'. Together they form a unique fingerprint.

Cite this