An efficient gesture based humanoid learning using wavelet descriptor and MFCC techniques
- PDF / 5,047,696 Bytes
- 20 Pages / 595.276 x 790.866 pts Page_size
- 72 Downloads / 139 Views
ORIGINAL ARTICLE
An efficient gesture based humanoid learning using wavelet descriptor and MFCC techniques Neha Baranwal1 • G. C. Nandi1
Received: 23 May 2015 / Accepted: 5 February 2016 Springer-Verlag Berlin Heidelberg 2016
Abstract Recognizing any gesture, pre-processing and feature extraction are the two major issues which we have solved by proposing a novel concept of Indian Sign Language (ISL) gesture recognition in which a combination of wavelet descriptor (WD) and Mel Sec Frequency Cepstral Coefficients (MFCC) feature extraction technique have been used. This combination is very effective against noise reduction and extraction of invariant features. Here we used WD for reducing dimensionality of the data and moment invariant point extraction of hand gestures. After that MFCC is used for finding the spectral envelope of an image frame. This spectral envelope quality is useful for recognizing hand gestures in complex environment by eliminating darkness present in each gesture. These feature vectors are then used for classifying a probe gestures using support vector machine (SVM) and K nearest neighbour classifiers. Performance of our proposed methodology has been tested on in house ISL datasets as well as on Sheffield Kinect gesture dataset. From experimental results we observed that WD with MFCC method provides high recognition rate as compare to other existing techniques [MFCC, orientation histogram (OH)]. Subsequently, ISL gestures have been transferred to a Humanoid HOAP-2 (humanoid open architecture platform) robot in Webots simulation platform. Then these gestures are imitated by HOAP-2 robot exactly in a same manner.
& Neha Baranwal [email protected] G. C. Nandi [email protected] 1
Robotics and AI Lab, Indian Institute of Information Technology, Allahabad, India
Keywords Computer vision DWT Indian sign language MFCC SVM
1 Introduction Nowadays robot can assist in various fields like doctors in his/her surgery during the critical operations, war field, household applications, etc. For such type of activity, interaction of human with robot is necessary. Many algorithms and many methodologies have already been evolved and many research works are currently running to make the robot/machine as intelligent as human beings. Both gesture and speech are the good ways of establishing a communication between human and robot [1, 2]. Gestures can be formed by hand and head movements or sometimes by full body movements which is mostly used by hearing impaired society. This society uses this sign language for establishing a communication with each other. In this work hand gestures have been taken into account. Previously gestures are recognized using data glove based gesture capturing techniques where sensors are used for capturing the joint angle values of hand. Using these angle values movements of hand gestures are identified. This method is not substantial for recognizing hand gestures therefore vision based gesture recognition technique have been evolved. This technique has the good capa
Data Loading...
