Measure tracking feature request

[marabuuu]

This is a feature request for implementing the following tracking features:

• velocity
• total_distance_travelled
• number_of_splits
• directional_change_rate

Best,
Mara

[haesleinhuepf]

Hi @somas193 ,

do you think this would be feasible? What we would need is a function that consumes a labels layer:

def compute_track_features(labels_layer:"napari.layers.Labels") -> "napari.layers.Labels":

And computes these features. It creates a new layer (maybe a track-ID labels layer?) where .features contain a pandas dataframe with the measurements listed above.

Happy to explain in detail in a meeting!

Thanks!

Best,
Robert

[haesleinhuepf]

@marabuuu : Do you have a mathematical definition of directional_change_rate by chance? A paper where it's used might also be helpful. Thanks!

[marabuuu]

Hi @haesleinhuepf ,

Here is the mathematical definition for mean directional change according to the TrackMate Documentation
In the TrackMate source code, there is a variable called TRACK_MEAN_DIRECTIONAL_CHANGE_RATE which is computed out of DIRECTIONAL_CHANGE_RATE.

I did not find a suiting paper though. If this one is too complicated, we can also leave it out. Just thought it might be cool.

[somas193]

Hi @somas193 ,

do you think this would be feasible? What we would need is a function that consumes a labels layer:

def compute_track_features(labels_layer:"napari.layers.Labels") -> "napari.layers.Labels":

And computes these features. It creates a new layer (maybe a track-ID labels layer?) where .features contain a pandas dataframe with the measurements listed above.

Happy to explain in detail in a meeting!

Thanks!

Best, Robert

@haesleinhuepf we can discuss it tomorrow some time if you have time.