Using Watershed Transform for Vision-based Two-Hand Occlusion in an Interactive AR Environment.

Saved in:

Bibliographic Details
Title:	Using Watershed Transform for Vision-based Two-Hand Occlusion in an Interactive AR Environment.
Authors:	Peng Peng Leim¹ pengpeng_leim89@hotmail.com, Guat Yew Tan¹ gytan@cs.usm.my, Kah Pin Ng¹ nkp10_mah017p@student.usm.my, Miin Huey Ang¹ mathamh@cs.usm.my
Source:	Annual International Conference on Information Technology & Applications. 2014, p190-194. 5p.
Subjects:	Augmented reality, Interactive model (Communication), Interactive computer systems, Interaction design (Human-computer interaction), Virtual design, Virtual communications
Abstract:	To achieve a natural interaction in augmented reality environment, we have suggested to use markerless vision-based two-handed gestures for the interaction; with an outstretched hand and a pointing hand used as virtual object registration plane and pointing device respectively. However, two-handed interaction always causes mutual occlusion which jeopardizes the hand gesture recognition. In this paper, we present a solution for two-hand occlusion by using watershed transform. The main idea is to start from a two-hand occlusion image in binary format, then form a grey-scale image based on the distance of each non-object pixel to object pixel. The watershed algorithm is applied to the negation of the grey scaled image to form watershed lines which separates the two hands. Fingertips are then identified and each hand is recognized based on number of fingertips on each hand. The outstretched hand is assumed to contain 5 fingertips and the pointing device contains less than 5 fingertips. An example of applying our result in hand and virtual object interaction is displayed at the end of the paper. [ABSTRACT FROM AUTHOR]
	Copyright of Annual International Conference on Information Technology & Applications is the property of Global Science & Technology Forum and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database:	Engineering Source

Description
Abstract:	To achieve a natural interaction in augmented reality environment, we have suggested to use markerless vision-based two-handed gestures for the interaction; with an outstretched hand and a pointing hand used as virtual object registration plane and pointing device respectively. However, two-handed interaction always causes mutual occlusion which jeopardizes the hand gesture recognition. In this paper, we present a solution for two-hand occlusion by using watershed transform. The main idea is to start from a two-hand occlusion image in binary format, then form a grey-scale image based on the distance of each non-object pixel to object pixel. The watershed algorithm is applied to the negation of the grey scaled image to form watershed lines which separates the two hands. Fingertips are then identified and each hand is recognized based on number of fingertips on each hand. The outstretched hand is assumed to contain 5 fingertips and the pointing device contains less than 5 fingertips. An example of applying our result in hand and virtual object interaction is displayed at the end of the paper. [ABSTRACT FROM AUTHOR]