filterTracks.m

Class: Function

Description: Finds all tracks with mean feature values (averaged across time) that exceed a user-specified value. User can specify whether tracks with mean feature values that fall below or above this threshold are removed.

Author: Oliver J. Meacock

filterTracks.m

function procTracks = filterTracks(procTracks,filterField,filterVal,filterDir)
%FILTERTRACKS removes tracks with mean feature values that exceed a
%user-specified threshold.
%
%   INPUTS:
%       -procTracks: The main track output of FAST's tracking module
%       -filterField: String specifying a feature field you wish to use as
%       a filter (e.g. 'vmag','sparefeat1' etc)
%       -filterVal: The value that should not be exceeded for a given track
%       to remain under consideration.
%       -filterDir: Whether you wish to remove tracks below (-1) or above
%       (+1) the specified threshold value.
%
%   OUTPUTS:
%       -procTracks: The set of processed tracks, with tracks that exceed
%       the specified threshold removed.
%
%   Author: Oliver J. Meacock, (c) 2021.
 
meanVals = arrayfun(@(x)(nanmean(x.(filterField))),procTracks);
 
switch filterDir
    case +1
        badInds = meanVals > filterVal;
    case -1
        badInds = meanVals < filterVal;
end
 
procTracks(badInds) = [];

Example usage:

%Extract and plot mean intensity of each object in channel 1
meanInts = arrayfun(@(x)(nanmean(x.channel_1_mean)),procTracks);
histogram(meanInts,'BinWidth',5)
hold on
 
%Remove all objects that have a channel 1 intensity that exceeds this threshold
procTracks = filterTracks(procTracks,'channel_1_mean',60,+1);
 
%Repeat intensity measurement and plotting
meanInts = arrayfun(@(x)(nanmean(x.channel_1_mean)),procTracks);
histogram(meanInts,'BinWidth',5)

Example output:

After filtering, all tracks with average channel 1 intensities greater than 60 have been removed: