Hand Gesture Recognition Names Utilizing Hidden Markov Model for Computer Visions Application

Abstract

This work focuses on advancing natural and intuitive human-computer interaction through the application of Hidden Markov Models (HMMs) in hand gesture recognition. The study addresses challenges in existing gesture recognition systems by implementing HMMs to capture temporal dynamics and diverse gestures. Integration with computer vision techniques enhances real-time processing, making the system adaptable to various environments. The methodology includes a literature review, a detailed implementation process involving video input, segmentation, morphological operations, hand tracking, and trajectory smoothing. The work successfully recognizes hand gestures in real-time video streams, showcasing applications in human-computer interaction, virtual reality, and gaming. The incorporation of the Baum-Welch re-estimation algorithm optimizes HMM parameters, leading to accurate recognition of specific names associated with hand gestures. Overall, the work contributes to the development of a robust and flexible framework for natural and intuitive gesture recognition.

References