Sign Language Recognition System to Aid Deaf-Dumb People Using Contourlet Transform
Communication between deaf-mute and a normal person is a challenging job. The Sign Language is a method to
communicate with deaf-dumb people. The goal of the work is to design and implement a real time Sign Language
Recognition system, to recognize 26 alphabets from Universal English Sign Language using MATLAB . The proposed
scheme consists of five modules: image acquisition, pre-processing and segmentation, feature extraction, sign recognition
and voice conversion. The Database is created and developed, for training and testing phases. The hand signs of the
alphabets are acquired, transformed using Contourlet Transform and features were extracted, then they are classified and
recognized using KNN classifier and are stored in a file, finally the file content is converted into audio using the
external - software program. This work enables deaf-mute people to communicate with non-signing people without the
demand of an interpreter.
Keywords- Sign Language, Feature Extraction, Sign Recognition, Contourlet Transform, Segmentation, Gestures.