A Visual Recognition of Static Hand Gestures in Indian Sign Language based on Kohonen Self-Organizing Map Algorithm

Indian Sign Language (ISL) or Indo-Pakistani Sign Language is possibly the prevalent sign language variety in South Asia used by at least several hundred deaf signers. It is different in the phonetics, grammar and syntax from other country’s sign languages. Since ISL got standardized only recently, there is very little research work that has happened in ISL recognition. Considering the challenges in ISL gesture recognition, a novel method for recognition of static signs of Indian sign language alphabets and numerals for Human Computer Interaction (HCI) has been proposed in this thesis work. The developed algorithm for the hand gesture recognition system in ISL formulates a vision-based approach, using the Two-Dimensional Discrete Cosine Transform (2D-DCT) for image compression and the Self-Organizing Map (SOM) or Kohonen Self Organizing Feature Map (SOFM) Neural Network for pattern recognition purpose, simulated in MATLAB. To design an efficient and user friendly hand gesture recognition system, a GUI model has been implemented. The main advantage of this algorithm is its high-speed processing capability and low computational requirements, in terms of both speed and memory utilization. KeywordsArtificial Neural Network, Hand Gesture Recognition, Human Computer Interaction (HCI), Indian Sign Language (ISL), Kohonen Self Organizing Feature Map (SOFM), Two-Dimensional Discrete Cosine Transform (2D-