Exploring Hand Gesture Recognition for Enhanced Human-Computer Interaction
S Kanagamalliga, P Vedasundara Vinayagam, Rajesh Kumar M, Rahul R · 2024
Hand gesture recognition is a key advancement in human-computer interaction, enabling real-time interpretation of hand movements from video. This technology is useful for applications like volume control, where it provides an intuitive, non-contact way to interact with devices. The main challenge is accurately recognizing gestures made by a single hand. This research introduces a new method that combines shape-based feature recognition with adaptive thresholding to enhance gesture accuracy and minimize errors. A Haar-cascade classifier is used to detect hand regions efficiently. A single camera captures user gestures, which are then processed in several steps: capturing the gesture, segmenting and identifying the hand in video frames, and recognizing the gesture based on its shape. The adaptive thresholding technique adjusts sensitivity based on hand size and distance from the camera, improving performance in different conditions. This method allows for accurate and efficient real-time hand detection. The goal is to create a system that can identify specific gestures and control devices without physical input devices like keyboards or mice, making interaction more convenient and accessible. The research demonstrates the design of this gesture detection system, highlighting the use of adaptive thresholding and its potential applications. The findings propose that such systems can considerably improve user experience by offering a user-friendly interface for controlling devices. Real-time gesture recognition provides a way to interact with computers through simple hand movements, with potential applications extending to sign language interpretation and other areas of human-computer interaction.