Machine Vision Inspection Systems, Image Processing, Concepts, Methodologies, and Applications

2Indian Sign Language Recognition Using Soft Computing Techniques

Ashok Kumar Sahoo¹*, Pradeepta Kumar Sarangi² and Parul Goyal³

¹ Department of Computer Science and Engineering, Graphic Era Hill University, Dehradun, India

² Chitkara University Institute of Engineering and Technology, Chitkara University, Punjab, India

³ School of Computing, Graphic Era Hill University, Dehradun, India

Abstract

Sign language recognition comes under the research dimension of pattern recognition. The work is to translate acquired images or videos either offline or online to corresponding words, numbers, or sentences representing meaning of the input sign. Here the work presented is recognition of Indian Sign Language. The application of this work is limited to Indian subcontinent where around 5% of the population is using Indian Sign Language to communicate with their external world. The direct pixel value, hierarchical centroid, local histogram features of image processing techniques are used in the experiments. The classifiers used here are k-Nearest Neighbor and Neural Network. The detailed work in this chapter is presented below.

Keywords: Indian sign language, feature extraction, histogram, hierarchical centroid, direct pixel value, naive Baÿes, k-nearest neighbor classifier, neural network classifier

2.1 Introduction

Sign language [1] is the communication medium for hearing-impaired people and hard hearing community. This is also helpful in communication between speaking and nonspeaking ...