Improved subject calibration

The new Nexus 2 labeler makes use of additional information available in Nexus 2 labeling skeleton templates (VSTs). This information is provided via the subject calibration workflow. The subject calibration workflow consists of incremental steps that enable you to balance the time taken for calibration against the overall extent of the information in the labeling skeleton (VSK), depending upon your specific labeling requirements.

The new Auto Initialize Labeling pipeline enables you to quickly calibrate a subject for simple labeling. It provides the following operations:

Operation

Description

Autolabel Static Frame

Attempts to label the subject when they are standing in a known pose. It matches the shape of the reconstruction to the shape of the subject template for that pose.

The labels are applied to the whole trajectory, not just the chosen frame. It works best when all the markers have been reconstructed and there are no additional markers or clutter in the scene.

Scale subject

Scales the labeling skeleton to be the same size as a labeled set of reconstructions on a particular frame. The scale factor is calculated assuming that the subject is in the same pose as was used for Autolabel Static Frame.

This scaling respects any constraints that you have specified in the labeling skeleton template. For example if your template specifies (ie a StaticParameter) that a segment is 100mm long, the segment will not be scaled. Only parameters are considered for scaling.

Marker-only Subject Calibration

Uses a single frame to update the marker locations for the labeling skeleton. This operation both fits the subject pose (that is, the pose that was defined when the labeling skeleton was created) to the labeled reconstructions and calibrates the marker locations.

Fitting the subject pose allows for the likely situation of your subject standing close to the template pose. Again, this operation respects any constraints specified in the labeling skeleton template and changes only parameters.

All of the operations make their calculations on a single frame, however, any labels that are assigned are applied to the whole trajectory.

To use automated labeling, you must follow a slightly different workflow from that previously used in Nexus.

The new Auto Initialize Labeling pipeline uses the supplied updated labeling skeleton templates files (VSTs) to reduce, or in many cases, eliminate the need for manual labeling. The new pipeline also works with custom VSTs that are built using the Labeling Template Builder in Nexus 2.x.

Requirements for custom labeling skeleton templates (VSTs)

To obtain the best results from the Auto Initialize Labeling pipeline, ensure your custom VSTs meet the following guidelines:

     Are built using the Labeling Template Builder in Nexus 2.x.

     Have at least three markers per segment.

     Use a repeatable base pose (used for initial labeling).

In addition, ensure that:

     A good ROM trial is used for the initial calibration.