Skeleton calibration operations in detail

Calibrating a Vicon labeling skeleton (VST/VSK) is the process that enables Nexus to recognize the subject and automatically determine which labels belong to the reconstructions. There is more than one process and multiple calibration options. A number of factors can affect which process and which options are best for your data/situation.

For a step-by-step introduction to the calibration workflow in which the operations described in this topic are used, see New workflow for Vicon Nexus 2.

The following types of trial can be used to calibrate a labeling skeleton:

❙     ROM trial This is a Range Of Motion trial. You can use a ROM trial to create a labeling skeleton for the subject. In a ROM trial, the subject completes a series of movements that exercises all of their joints. To produce the best calibration (and auto-labeling), the subject moves all of their joints through as full a joint range as possible. For the best results, ensure that each joint is moved through a range that represents what the subject is likely to do during collection. Full ROM calibrations provide the most information of any of the methods and therefore often provide the best labeling performance outcomes. For this reason Vicon recommends using a full ROM calibration whenever possible.

❙     Static trial This is a short trial where the subject stands in a base pose. A base pose is a static pose that is used when you first create the VST file. This is often a T-pose or 'motorbike' pose (see Capture and process a ROM trial). As with the ROM trial, the purpose of this trial is subject setup. With a well-designed VST, a static trial can be used to calibrate a labeling skeleton. You do this by using the Auto Initialize Labeling pipeline for Plug-in Gait. This operation requires the least processing time as it operates on only a single frame of data. However, static calibration provides less information than functional calibration, for the same reason. This type of calibration trial can be very helpful if:

–           A subject’s ability to perform a ROM trial is limited.

–           Time is factor.

❙     Dynamic trial This is a trial in which the subject performs the activity that is being studied. As such this trial is similar to a ROM trial, but focuses on only joint movements that are expected and specific to the task/activity being performed. In a gait laboratory, an example would be using a walking trial for this calibration. These trials are not normally used for an initial or full subject setup, but can be used in combination with some operations to improve labeling. Dynamic activity-matched trials may be of benefit if the movement of the subject during collection is not similar enough to what is captured when using a Static or ROM trial.

Subject-specific information is what enables a skeleton labeling template (VST) to be converted to a subject-specific labeling skeleton (VSK). All of the skeleton calibration operations make changes to the labeling skeleton, as can be seen inside the VSK file. For VST version 3 files, the attributes that are modified by at least one of the existing skeleton calibration operations are:

❙     Parameters. These control the pose of joints and the position of markers in the parent segment coordinate frame. A single parameter can be applied to both a segment and a marker or any combination of markers and segments. The calibration operations can change the value stored in the parameter.

❙     Segments. Bone lengths can be changed due to the parameters changing. The VST format doesn't have a concept of bone length. Bone lengths are inferred from the pose transformation between a pair of joints. This is made up of a pre- and a post-transformation. It is quite common for this transformation to have only one parameter that can be changed: this parameter is often named SomeBoneLength. The calibrated values are written to the VSK but are not reloaded on VSK import.

❙     Joints. Various attributes on the joint can be changed by the calibration operations. The mean, covariance, range-center and range can all be calculated from data.

❙     Targets. Target (marker) mean and covariance can be calculated from data.

Comparison of skeleton calibration operations

Deciding which of the different skeleton calibration operations is best in your situation depends on a number of considerations including trial type, processing time, and desired level of labeling accuracy.

The following descriptions cover the various operations that use the same underlying skeleton calibration algorithm.

To use any of the operations, observe the following preconditions:

❙     A fully labeled trial (ROM, static, or dynamic) must exist.

❙     The trial must contain only raw reconstructions; leave any gaps unfilled.

Tip:          Unlabeled reconstructions (obviously) have no influence on the operations.

❙     Functional Skeleton Calibration operation

❙     Functional Skeleton Calibration - Markers Only operation

❙     Static Skeleton Calibration operation

❙     Static Skeleton Calibration - Markers Only operation

❙     Calculate Skeleton Joint & Marker Statistics operation

The operations change the following pieces of information in the skeleton

For information on how to use the operations in common Nexus workflows, see Subject calibration workflows.