A mathematical interpretation of human body movement

Amanzadeh, Mohammad Mahdi

Permalink

https://hdl.handle.net/2142/115388 Copy

Description

Title

A mathematical interpretation of human body movement

Author(s)

Amanzadeh, Mohammad Mahdi

Issue Date

2022-04-13

Director of Research (if dissertation) or Advisor (if thesis)

Francis, George K

Doctoral Committee Chair(s)

Francis, George K

Committee Member(s)

• Ruecker, Stan
• Toenjes, John M
• Carpenter, Bruce

Department of Study

Illinois Informatics Institute

Discipline

Informatics

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

• Human-Computer Interaction
• Human Body Movement
• Application Programming Interface
• Computer Graphics
• Computer Animation

Abstract

The overarching theme of this research project is to build a machine-accessible and mathematically precise abstraction of human body movement through experiments in programming an articulated puppet that imitates the human body structure and mobility within a Real-Time Interactive Computer Animation (RTICA). Human movement has been the subject of research in science, art, and engineering, to design mental models, practices, and technology for observing, exercising, performing, and interpreting movements. With the advent of motion capture technology, it is possible to design software and hardware interfaces for human-machine interaction through movement, such as games and virtual environments for rehabilitation, education, and entertainment. A problem in designing application programs for the human body is the lack of a machine-accessible movement notation system for transcribing human movement and its properties, such as the expressed intention and mood. Such a notation system serves as an interface that encodes the shape and dynamics of a performer’s movement into sequences of parameters that can be interpreted and reconstructed by a computer program, while the parametrization remains invariant to the geometric structure, position, and orientation of the performer’s body, as observed in the space of the motion capture camera. Through experiments in programming a real-time interactive computer animation, we explore the instances of a training-learning paradigm in which a teacher demonstrates a movement to a student who tries to learn the performance by imitating the teacher. The teacher and the student RTICAs are two instances of an articulated puppet that mimics the mechanical linkage of the human body by a hierarchy of Euclidean transformations that control the oscillation of each puppet limb in the coordinate system of the limb’s preceding joint. The teacher RTICA represents a clockwork automaton that can be programmed to move using an animation program, which specifies the puppet’s keypose at certain intervals of the movement. We extend the design of the teacher RTICA with an abstraction of the brain’s mirror system, to construct the student RTICA, which represents an interactive automaton that can imitate the motion data of a human movement to parametrize the performance into an animation program. The student RTICA serves as an Application Programming Interface (API) for human body movement, which permits the interaction between a performer and the controller system of an application program for the human body. The API encodes the motion data of a performer’s movement by the sequences of keypose data that parametrize the oscillation of the individual limbs, which are then interpolated and superimposed over the graph structure of the puppet’s mechanical linkage to reconstruct the movement.

Graduation Semester

2022-05

Type of Resource

Thesis

Copyright and License Information

Copyright 2022 Mohammad Amanzadeh

Owning Collections