A sign-logo image search & combination system by analyzing color and shape features

Yuki Toyoshima, Yasuhiro Hayashi, Yasushi Kiyoki

Research output: Contribution to conferencePaperpeer-review

3 Citations (Scopus)

Abstract

This paper presents a sign-logo search and combination system by analyzing the color and shape features of sign-logos. The objective of this system is to find a global way of communicating by combining sign-logo images, which are created by combining color and shape features, to express important messages with visualizations. This system would contribute to communication between foreign people who uses different languages, and to share the same interpretations during collaborating scenes. To realize this contribution, this system analyses the color and shape features of the image input by a user, also divides the shape into frames and objects, which consists of color layers. The frame stands for the context of the message, and the objects stand for the target of the message. The colors used in the sign-logos tend to implicitly express the importance of the message. The results of several experimentations for evaluating effectiveness of our system are shown in this paper.

Original languageEnglish
Pages313-320
Number of pages8
DOIs
Publication statusPublished - 2013
Externally publishedYes
EventIASTED International Conference on Signal Processing, Pattern Recognition and Applications, SPPRA 2013 - Innsbruck, Austria
Duration: 2013 Feb 122013 Feb 14

Other

OtherIASTED International Conference on Signal Processing, Pattern Recognition and Applications, SPPRA 2013
Country/TerritoryAustria
CityInnsbruck
Period13/2/1213/2/14

Keywords

  • Content-based image retrieval
  • Image database
  • Image-query creation
  • Sign-logo
  • Visual communication

ASJC Scopus subject areas

  • Computer Vision and Pattern Recognition
  • Signal Processing

Fingerprint

Dive into the research topics of 'A sign-logo image search & combination system by analyzing color and shape features'. Together they form a unique fingerprint.

Cite this