Segmentation of the character string pointed out by a finger tip from a pair of color images

Shinj Tsuruoka, Makoto Nishikawa, Xinkai Chen, Muneaki Ishida

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

Abstract

We propose a Visual Pointing Out System (VPOS) using a finger tip to indicate a character string from a pair of color images to a computer or a robot. The image pair consists of an image with character strings and a hand, and an image without a hand. This system includes the segmentation of the string regions, the extraction of the hand region, the specification of the finger tip and the extraction of the string region indicated by the finger tip. The algorithm employs morphology (dilation and erosion), a skin color distribution table and geometrical analysis. We have evaluated the effectiveness of this method by computer experiments. Results show that the algorithm can extract both the image of a finger tip from an extended finger, and the character string region indicated by the finger tip. It will be useful and practical as a man machine interface technique.

Original languageEnglish
Title of host publicationProceedings - 3rd Asia Pacific Computer Human Interaction, APCHI 1998
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages310-316
Number of pages7
ISBN (Print)0818683473, 9780818683473
DOIs
Publication statusPublished - 1998 Jan 1
Externally publishedYes
Event3rd Asia Pacific Computer Human Interaction, APCHI 1998 - Hayama-machi, Kanagawa, Japan
Duration: 1998 Jul 151998 Jul 17

Other

Other3rd Asia Pacific Computer Human Interaction, APCHI 1998
CountryJapan
CityHayama-machi, Kanagawa
Period98/7/1598/7/17

Keywords

  • 3-D interface
  • morphology
  • multi-media
  • multi-modal communications
  • robot vision

ASJC Scopus subject areas

  • Human-Computer Interaction
  • Computer Graphics and Computer-Aided Design

Fingerprint Dive into the research topics of 'Segmentation of the character string pointed out by a finger tip from a pair of color images'. Together they form a unique fingerprint.

Cite this