<![CDATA[The need for effective and affordable post-stroke rehabilitation technology has driven the development of a finger gesture detection system based on YOLOv8 implemented on the Jetson Xavier NX. The system is designed to recognize hand gestures—including open hand, closed hand, and thumb touching other fingers—in real time using a camera as the image input. The detection results are then transmitted to an ESP32 microcontroller to control LEDs as visual indicators. The research process involves hardware and software design, model training using datasets from Roboflow, and system testing on the Jetson Xavier NX. Experimental results show that the system can detect gestures with high accuracy and fast response. These findings demonstrate the potential of computer vision based on edge computing as a foundation for assistive technologies in both medical rehabilitation and human–computer interaction.]]>
Copyrights © 2026
