Radially Defined Local Binary Patterns for Hand Gesture Recognition

Hand gestures are extensively used in human nonverbal communication by hearing impaired and speech impaired people. To make communication between the hearing impaired and normal people simple and efficient, it is necessary that this process be automated. Number of techniques have been developed for automatic HGR. The usefulness of a method is analyzed based on various parameters like computational complexity, memory requirement, time required, etc. This paper proposes a new algorithm for static hand segmentation from complex background and for feature extraction for hand gesture recognition, considering the need of faster and simpler algorithms giving high recognition efficiency.