Three Dimensional Gesture Recognition Using PCA of Stereo Images and Modified Matching Algorithm

This paper describes 3-dimensional (3D) gesture recognition using principal component analysis. Existing 2-dimensional gesture recognition systems have shortcomings such as limitation of motion. In order to solve this problem, motion recognition systems using 3D information were proposed. As vision-...

Full description

Saved in:
Bibliographic Details
Main Authors: Jong-Min Kim, Maeng-Kyu Song
Format: Conference Proceeding
Language:English
Subjects:
Biological system modeling
Data mining
Feature extraction
Head
Image recognition
Mathematical model
Motion analysis
Principal component analysis
Skin
Virtual reality
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This paper describes 3-dimensional (3D) gesture recognition using principal component analysis. Existing 2-dimensional gesture recognition systems have shortcomings such as limitation of motion. In order to solve this problem, motion recognition systems using 3D information were proposed. As vision-based 3D information has a number of dimensions as well as a lot of errors, however, it was difficult for those systems to find out consistent characteristics. In this paper, we describe a method of modeling and analyzing gestures using principal component analysis. This method helps reduce the influences of errors that 3D data might have and achieve the effect of dimension reduction. We also propose a matching algorithm modified to reduce the motion limitation in the model-based motion recognition system, and present examples of using the result of motion recognition as the interface for 3D action games.
DOI:10.1109/FSKD.2008.610