제목

The Control of Robot through Recognition of Human Arm Gesture using Mean Shift Algorithm

저자명

Jaepil LEE, Youngjoon HAN, Hernsoo HAHN

국문초록

This paper proposes the recognition of a human arm gesture to control robot’s arm effectively. The existing algorithms for arm gesture use and learn single information to recognize arms while these methods reduce computing speed and have difficulty in

interacting with tracking arms. Also, these could affect robot’s movement which causes delays.[1] Therefore, this paper uses three different methods for complementing problems; Haar-like feature algorithm to detect faces and shoulders, Mean-shift algorithm based

on hand’s color to detect and track hands, and Kinematics method to track arms and joints. In addition, divide an image into left and right area to track arms precisely in case of missing arms while tracking. Information provided through these methods is used to

recognize human arm gesture more accurately in order to deliver information to robot and to realize.

원문 수록처

Jounal of Measurement Science and Instrumentation

자료유형

International Journal