ZEISS Knowledge Base

Performing manual tracking on existing segments»

This guide explains how to perform object tracking using existing segments (previously detected).

The segments belonging to the track must be manually selected before applying the pipeline to generate the track.

Detecting the objects

Run the appropriate pipeline to detect the segments over time.
If required, apply the pre-processing steps followed by the Segmentation tools.

Any segmentation operator can be used to generate the segments. Blob Finder, Watershed, Machine learning tools are available and applicable accordingly to the dataset features.
The segments can also be generated importing labeled images.
The segments must be detected over time. Therefore, all the time points should show the found objects.

The tracking pipeline must be created. Here below an example:

Input ROI

Import Document Objects

Tag Filter: Sets the TAG(s) previously created (Track#1 and so on) from the manual object's selection procedure.
Tracking: Set the parameters to track the selected segments according to the object's behavior.
Store Objects: Set the data to be transferred to the objects table.

The Tag Filter can import multiple TAGS, each one will generate a single track.

The track is computed and the data will be transfered to the object table.

The needed features can be selected by using the Feature Columns tool.

To modify the tracks, open the Track Editor.
The tracks are listed on the left panel. The tracked objects belonging to the specific track are shown on the right panel.
Click on the track.
Select the time point and the segment to be edited. The related frame is shown in the viewing area. The track and the segment are highlighted.
Right-click on the selected time point.
A pop up menu is displayed.

The track can be splitted and merged with another one. The segments in the track path can also be deleted.