Logo

Static Hand Gesture Recognition Using Local Gabor Filter

Gupta, Shikha and Jaafar , jafreezal and Wan Ahmad, Wan Fatimah (2012) Static Hand Gesture Recognition Using Local Gabor Filter. [Citation Index Journal]

[img]
Preview
PDF (IRIS2012)
421Kb

Abstract

With the increasing demand of Human-Computer interaction it is expected that in future human computer interfaces will provide more convenient, natural and comfortable communication same like human-human interaction. Usage of static gestures in our daily life to convey certain meaning leads to recognition of static hand gestures as an important aspect in HCI. Recognition systems involve various processes such as feature extraction, features reduction and classification. Gabor filter is being used for feature extraction since long time in image processing due to its remarkable mathematical and biological properties. Gabor filters have high dimensionality so in the proposed method instead of using 40 Gabor filters only 15 local Gabor filters have been used. The main objective behind using 15 Gabor filter responses is to reduce the complexity with better accuracy. In the proposed method after using local Gabor filters the features are being reduced by PCA to overcome small sample size problem. So that can use LDA which is most popular linear projection technique for feature extraction and feature reduction. This combination of PCA and LDA has been used in many applications such as face recognition for security purpose. But here in this paper this is being used with Gabor filter for hand gesture recognition. Till now in hand gesture recognition the main focus is on Gabor filter with PCA but here in this paper will use PCA and LDA for hand gesture recognition. The use of local Gabor filter helps in reducing the redundant data as instead of using 40 filters here in this paper 15 filters are being used which is a subset of the global filters parameters. Classification of the gestures as per their classes will be done with the help of one against one multiclass SVM.

Item Type:Citation Index Journal
Impact Factor:1.65
Subjects:T Technology > T Technology (General)
Academic Subject One:Hand gesture
Departments / MOR / COE:Departments > Computer Information Sciences
ID Code:11076
Deposited By: Assoc Prof Dr Wan Fatimah Wan Ahmad
Deposited On:16 Feb 2014 23:41
Last Modified:16 Feb 2014 23:41

Repository Staff Only: item control page

Document Downloads

More statistics for this item...