Research on Algorithm and Model of Hand Gestures Recognition Based on HMM

Human computer interaction is one of the key points in the competition of information industry in the world, all countries in the world put the human-computer interaction as a key technology to study. Butler Lampson, ACM Turing Award winner in 1992 and Microsoft Research Institute chief software engineer pointed out that the computer has three functions in the “21st century computing research” report. The first is simulation; the second is that the computer can help people to communicate; the third is interaction, that is, to communicate with the real world. Human-computer interaction is an important field of computer research, and hand gestures recognition is a key technology in this field. The key of gesture recognition is the feature extraction and the establishment of hand recognition model. It can accurately identify the various kinds of deformation. HMM method has a flexible and efficient training and recognition algorithm, if the system needs to add a new gesture, just need to train the gesture of the sample set can be; If a gesture is not needed, just delete the corresponding HMM algorithm of the gesture, HMM has a strong expansion. Compared with DTW and other methods, HMM in speech recognition, gesture recognition, the recognition effect is better. In this paper, the HMM algorithm is used to identify the typical gestures, got very good recognition effect.