Evolutionary motion inverse kinematics
Date
2007
Authors
Millard, John Terence
Editors
Advisors
Journal Title
Journal ISSN
Volume Title
Type:
Conference paper
Citation
Proceedings of IEEE Congress on Evolutionary Computation, 25-28 September, 2007: pp.3671-3678
Statement of Responsibility
Conference Name
IEEE Congress on Evolutionary Computation (2007 : Singapore)
Abstract
In disciplines, such as robotics and computer graphics (CG) or industries such as, game development, various tasks involve the manipulation of articulated structures in interesting and complex ways. This may involve controlling robotic manipulators or perhaps posing the skeleton of a virtual human. When performed manually, the task of rotating each individual joint in order to produce a desired pose can be tedious and time consuming. The process of rotating each joint independently is known as forward kinematics (FK). One way to simplify this process is to automatically calculate the angles required for a given joint (known as the end-effector) to reach a goal position - this process is known as inverse kinematics (IK). In this paper an evolutionary algorithm (EA) is proposed as a solution to inverse kinematics. Several extensions are also proposed to allow extra constraints to be included, such as obstacle avoidance and angular joint limits. This approach is called Evolutionary Motion IK (E-MOTION IK).
School/Discipline
School of Computer Science
Dissertation Note
Provenance
Description
Access Status
Rights
© 2008 IEEE – All Rights Reserved