Sign Language Recognition Using Hilbert Curve Feat

Authors: Amira Ragab Maher Ahmed SiuCheung Chau

Publish Date: 2013/6/26

Volume: , Issue: , Pages: 143-151

PDF Link

Abstract

In order to design an efficient visionbased gesture recognition system the images of the hand gestures need to be represented in an accurate way This paper presents a new method for representing hands’ images based on the Hilbert spacefilling curve The method first segments the hand and then applies a Hilbert spacefilling curve to extract a feature vector Afterwards a classifier such as Support Vector Machine SVM and Random Forest RF are used to classify the gestures The Hilbert curve representation is chosen in this work since it has proved its efficiency in representing shapes with uniform background due to its localizationpreserving property of pixels Moreover it is also known of being invariant to translation scaling and stretching The results reveal the efficiency and suitability of the proposed approach compared to that of other works in real time

Keywords: