Improving Dynamic Bounding Box using Skeleton Keypoints for Hand Pose Estimation

Autor: Arief Syaichu Rohman, Egi Hidayat, Carmadi Machbub, Ignatius Prasetya Dwi Wibawa
Rok vydání: 2020
Předmět:
Zdroj: 2020 6th International Conference on Interactive Digital Media (ICIDM).
DOI: 10.1109/icidm51048.2020.9339667
Popis: Human-Computer Interaction (HCI) studies the design and implementation of computer technology, focused on the interaction between users and machines or computers. One of human-computer interface is using visual-based communication, which is the most widespread area in HCI research. In this paper, we discuss dynamic bounding box that follows the movement of human hand skeleton key points using a single RGB camera. By using wrist hand joint skeleton as setpoint, dynamic bounding box can move according to the hand movement, and the size of the bounding box will scale up automatically according to the hand size and its position from the camera. For hand pose recognition experiment results, the classification accuracy for hand pose estimation using dynamic bounding box performs a good result compare with using static bounding box.
Databáze: OpenAIRE