Joint Center Marking tool

Project taken

 

This project is already taken

 

 

Background

Character animation in video games and other virtual environments is often performed thanks to a motion capture system and an inverse kinematics (IK) algorithm. We own a set of IK tools that make use, among other, of the center of rotation of some body joints. These center of rotations are currently being estimated using a rather long calibration step and cannot be double checked efficiently afterwards.

hands mocap body

Project idea

This project will be divided in two related, yet distinct parts. The first project part is supposed to be finished by mid-semester.

First, we are going to introduce a new physical calibration tool used to indicate the location of a joint’s center of rotation, such as a clamp or a stick equipped with markers. The goal is to provide a set of software and hardware tools to help during the calibration process.

We will then be looking for a way to visually validate that the calibration process went well. One option would be to track a camera and render two images using the same viewpoint, one in the real world and one in the virtual environment. By overlaying these two images we will be able to see if a mistake was made during calibration.

Goals

–    Understand our current calibration tool and the file format it uses.
–    Implement a pointing or clamping prototype and evaluate its performances.
–    Introduce a visual calibration quality assessment tool (e.g. tracked camera).
–    Provide a polished, easy to use, and expandable interface.

Requirements

  • Programming: OOP and Unity scripting in C#
  • Computer Graphics principles: Rendering, transformation matrices
  • 3D geometry and quaternions: Vectors, cross products, rotations