Abstract: Sign language recognition is an important area of research with uses in a range of industries, including education, communication, and human-computer interaction. Trajectory-based human hand tracking is a popular technique for sign language recognition because it allows capturing dynamic hand movement during signing. This paper presents an overview of different trajectory-based human hand tracking methods, including template matching, feature-based tracking, and model-based tracking. This framework aims to accurately identify various one-handed dynamic isolated characters and interpret their intended meaning. Emphasis is placed on designing a system that can effectively recognize and interpret a wide variety of sign gestures, enabling effective communication and understanding between sign language users and individuals who do not know sign language.
Keywords: pre-processing, feature extraction, recognition.
|
DOI:
10.17148/IJARCCE.2023.12624
