The Maximum Distance Moved smoothing method

Aim

To remove outliers in the track, due to erratic detection of objects within the arena. For example, in a DanioVision experiment, if detection is not optimal, the margin of the well is sometimes detected as the fish.

 

important  If you see many errors, adjust the Arena Settings and/or the Detection Settings.

Procedure

1.Choose Acquisition > Track Smoothing Profile > Open.

2.Select the Maximum Distance Moved check box and click the Edit button.

 

3.Enter the threshold value of distance moved, above which the sample will be set to missing. If the distance moved from sample 1 to 2 is longer than the threshold, sample 2 is set to missing.

4.Click OK.

5.Plot the tracks to see the result.

Notes

▪tip  To know which value of distance to use as a threshold, plot the values of distance moved in the Integrated Visualization, and look at the spikes corresponding to the outliers. Take a value well below the value of the spikes. Be aware that a threshold being too low can remove good data.

 

Result (compare this with the picture at the top):

 

▪The Maximum Distance Moved method is applied after data editing, and after the Smoothing (Lowess) and the Minimal Distance Moved methods (when selected). The smoothed data, not the original data, are used for analysis and for export. Applying the Maximum Distance Moved method does not change the original data.