Efficient big image data retrieval using clustering index and parallel computation

Ja Hwung Su, Chu Yu Chin, Jyun Yu Li, Vincent S. Tseng

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

2 Scopus citations

Abstract

Image data has grown rapidly because of advances on photo capturing devices. In traditional, because the image data has not been huge, most past studies focused on the effectiveness improvement. However, accessing the images from a huge amount of image data needs a large cost. Hence, how to perform efficient image retrieval has been a hot topic in the last few decades. To this end, in this paper, we propose efficient big image data retrieval by using clustering index and parallel computation. In the offline stage, the images are grouped into a number of clusters. In the online stage, the relevant images to the query image are retrieved by a level-wise search. Our intent is to conduct a more efficient image retrieval method in comparison with traditional methods but keep the same effectiveness still. In the experiments, four types of retrieval are compared and our proposed parallelized image data retrieval is much faster than the other compared methods under the very close accuracies.

Original languageEnglish
Title of host publicationProceedings - 2017 IEEE 8th International Conference on Awareness Science and Technology, iCAST 2017
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages182-187
Number of pages6
ISBN (Electronic)9781538629659
DOIs
StatePublished - 1 Jul 2017
Event8th IEEE International Conference on Awareness Science and Technology, iCAST 2017 - Taichung, Taiwan
Duration: 8 Nov 201710 Nov 2017

Publication series

NameProceedings - 2017 IEEE 8th International Conference on Awareness Science and Technology, iCAST 2017
Volume2018-January

Conference

Conference8th IEEE International Conference on Awareness Science and Technology, iCAST 2017
CountryTaiwan
CityTaichung
Period8/11/1710/11/17

Keywords

  • big image data
  • clustering index
  • image search
  • parallel computation
  • visual-based

Fingerprint Dive into the research topics of 'Efficient big image data retrieval using clustering index and parallel computation'. Together they form a unique fingerprint.

Cite this