IJCSI International Journal of Computer Science Issues, Vol. 8, Issue 6, No 1, November 2011 ISSN (Online): 1694-0814 www.IJCSI.org 391 Dynamic Gesture Recognition Using Hidden Markov Model in Static Background Malvika Bansal 1 , Shivin Saxena 2 , Devendra Desale 3 and Dnyaneshwar Jadhav 4 1 Department of Computer Engineering, MES College Of Engineering Pune University, Maharashtra, India 2 Department of Computer Engineering, MES College Of Engineering Pune University, Maharashtra, India 3 Department of Computer Engineering, MES College Of Engineering Pune University, Maharashtra, India 4 Department of Computer Engineering, MES College Of Engineering Pune University, Maharashtra, India Abstract Human Computer Interaction is a challenging endeavour. Being able to communicate with your computer (or robot) just as we humans interact with one another has been the prime objective of HCI research since the last two decades. A number of devices have been invented, each bringing with it a new aspect of interaction. Much work has gone into Speech and Gesture Recognition to develop an approach that would allow users to interact with their system by simple using their voice or simple intuitive gestures as against sitting in front of the computer and using a mouse or keyboard. Natural Interaction must be fast, convenient and reliable. In our project, we intend to develop one such natural interaction interface, one that can recognize hand gesture movements in real time using HMM but by using Computer Vision instead of sensory gloves. Keywords: HCI, Dynamic, Gesture Recognition, Skin colour detection, Computer Vision, HMM 1. Introduction Technology has come a long way from computers running on vacuum tubes to semi-conductor chips and finally, the super computers of today’s era. In the earlier stages of computing inventions, a keyboard was the only means of communication to interact with the computer (after magnetic tapes and punch cards became obsolete). The mouse brought with it a revolution and an entirely new dimension to Human Computer Interaction, or popularly abbreviated as HCI. Many inventions followed such as the light pen, tablets, digitizers and more recently the “Space Mouse” each bringing with it an innovative style of interacting with the computer and opening new possibilities and dimensions of interaction. Thus, Human Computer Interaction is not just a necessity but a challenging endeavour to push current boundaries of interaction. Speech Recognition and Synthesis has been a prominent domain of research over the last decade. Even more prominent has been interaction by the use of simple intuitive Hand Gestures. Sign language is a common form of communication between auditory handicapped people. It can further be employed to communicate with a robot or any computer. Imagine sitting on the couch and operation your computer from a distance with your voice or just simple day-to-day hand movements. It would not only eliminate the need to actually physically touch your mouse and keyboard unless absolutely necessary, but will also be so much more convenient and quick. This is the power of Human Computer Interaction. Both static and dynamic hand gesture recognition is a challenging aspect of HCI. Gesture Recognition can be implemented by one of the two methods, one, by wearing sensory gloves on one’s hand or second, by with the aid of Computer Vision (CV). For glove based techniques, it mainly utilizes sensory gloves to measure the angles and spatial positions of a hand and fingers. Referring to one of the papers based on this field, we came across an approach in gesture recognition that used a sensing glove with 6 embedded accelerometers. It could recognize 28 static hand gestures and the computation time was about 1 characters/second. However, the proposed algorithm was not efficient enough to be applied in real time. Although the former is a powerful technique it’s not really a natural way to interact with the computer because one has to continually wear the gloves. Natural HCI should be glove free, fast, reliable and convenient. A sensory glove would not be a convenient option for daily usage.