A segment is a rigid object defined by a local coordinate system and a length.

A segment is specified by a proximal endpoint, a distal endpoint, and an orientation.

In Visual3D, the origin of the local coordinate system is at the proximal end of the segment.

The use of the term proximal and distal segment endpoints is an unfortunate legacy expression in Visual3D because it isn't necessarily true that the proximal end of the segment corresponds to the anatomical proximal end of the segment, although this is often the case.

The orientation of the segment coordinate system is typically aligned with the anatomical axes, but it isn't necessarily true that it be aligned in this way. There are many reasons that a segment might be created using axes not aligned with the segment.

Visual3D assumes that the local coordinate system is capable of moving in space, so it is assumed that a sensor is attached to the segment to track this movement. The motion capture system (regardless of the type) tracks the position and/or orientation of the sensor and Visual3D computes the pose (position and orientation) of the segment from the transformation between sensor coordinates, segment coordinates, and joint constraints. The term sensor is used very loosely here because Visual3D actually assumes that markers are attached rigidly to a segment, or that a segment is connected to other segments by a “joint”. Visual3D will not allow a segment to be defined that doesn't not include tracking markers.

If physical markers are not attached to a segment because the segment is tracked using an inertial sensor, electromagnetic sensor, etc., Visual3D creates virtual markers to represent the output of the sensors. This is useful because it allows the output of a sensor to be visualized (e.g. by three non-colinear markers) and it means that all segment tracking is computed consistently regardless of the sensor used.