<![CDATA[Gesture recognition technology enables computers and digital devices to detect, understand, and interpret human body movements through image processing techniques. This technology has significant potential to facilitate communication between individuals with hearing impairments and those without, thereby improving interaction and mutual understanding. However, the accuracy of gesture recognition systems is often influenced by variations in the distances between hand landmark points, which can introduce instability and reduce recognition performance. To address this issue, this study proposes a polynomial regression-based approach to stabilize distance measurements between hand landmarks in gesture recognition tasks. The proposed method calculates and normalizes landmark distances using polynomial regression to minimize measurement fluctuations and improve recognition accuracy. The system is implemented using the MediaPipe framework for real-time hand detection and tracking, while OpenCV is utilized for video processing and management. Experimental results demonstrate that the proposed approach significantly enhances the stability and accuracy of gesture detection. The developed system successfully recognizes hand gestures representing the letters A through F with an average accuracy exceeding 98.3%. Furthermore, the application of polynomial regression effectively reduces noise in landmark data, contributing to more reliable and accurate gesture recognition performance.]]>
Copyrights © 2024
