Performing Fully Automated Motion Tracking using ROI Correlation

ROI-based motion tracking can be used to track objects when image segmentation is difficult or not possible. ROI-based tracking uses Fourier cross-correlation based on the ROI you draw around the object of interest as it appears on the first frame of the sequence in order to find the same object on subsequent frames.

To perform motion tracking using ROI correlation:

Open the sequence file containing the sequence of frames documenting the movement of your object of interest.

On the Count/Size tab, click Objects in the Tracking group.

The Motion Tracking ribbon is displayed.

Use the Learning tool to discover the approximate Motion Type, Velocity Limit, and Acceleration Limit parameters for your object of interest.  These values define the search radius for detecting the object of interested from frame to frame.

Click on the down-arrow of the Auto All tool and select Learning.

The Tracking Message dialog is displayed.

Click on the object of interest in the current frame of the sequence.

A track marker is placed in the sequence and the next frame is displayed.

Click on the object of interest in the next frame.

Another track marker is placed.

Repeat b and c for a representative sample of frames.

Click Close Track in the Tracking Message dialog.

A message box appears detailing the calculated Motion Type, Velocity Limit, and Acceleration for the object give the representative sample of frames you provided.

These values will be used by Image-Pro to auto-detect your object of interest across all frames of the sequence.

Click on the down-arrow of the ROI tool in the ROI Based Tracking group, and select the best tool for drawing an ROI around the object of interest. For example, if the object is circular, select the circle tool.

Go back to the first frame of the sequence and draw an ROI around the object of interest.

Click the Match tool.

The area around the current ROI position expanded by the search radius on the following frame will be analyzed to find the best-match position of the object. This process will be repeated for all active frames of the sequence. Then the tracking module will create a new track, using the  central positions of the AOIs.

 

Learn more about all features in Count/Size