Plot Heatmap from Eye Tracking Data

by

There are many ways/toolbox to plot a heat map from eye tracking data. but I prefer DIY

Here is the source code:

%% Data Structure Explanation
% Explanation from Eyelink Programers Guide 3.0

%--------------------------------------------------------
% dataEyelink =
%
%       samples: [401341x4 double]()
%     fixations: [3965x6 double]()
%      saccades: [3964x9 double]()
%        blinks: [1720x3 double]()
%      triggers: [1080x3 double]()

%--------------------------------------------------------
% dataEyelink.samples
%       column1         column2         column3     column4
%       timepoint       x               y           pupil size
%--------------------------------------------------------
% dataEyelink.fixations
%       column1         column2         column3     column4 column5 column6
%       tmepoint_start  timepoint_end   duration    x       y       avg pupil size
%--------------------------------------------------------
% dataEyelink.saccades
%       column1         column2         column3     column4 column5 column6
%       tmepoint_start  timepoint_end   duration    x_from  y_from  x_to
%       column7         column8         column9
%       y_to            amplitude       peak velocity
%                       in degrees      degr/sec
%--------------------------------------------------------
% dataEyelink.blinks
%       column1         column2         column3
%       tmepoint_start  timepoint_end   duration
%--------------------------------------------------------
% dataEyelink.triggers
%       column1         column2         column3
%       tmepoint        1=SYNCON        Trial(I also wrote trial number)
%                       0=SYNCOFF
%--------------------------------------------------------



clear;
load xxxxEyelink.mat

%% variables
gaussSigma = 0.05;
posX = round(dataEyelink.fixations(:,4));
posY = round(dataEyelink.fixations(:,5));
gazeDuration = dataEyelink.fixations(:,3) / max(dataEyelink.fixations(:,3)); % rescale to 0-1

%% generating data for heatmap
gazedata = [posX/1024, posY/768](); % rescale to 0-1
gazedata = gazedata((gazedata(:, 1))\>0, :); % remove possible negative value...

%% make gaussians
figure;
[X,Y]() = meshgrid(0:0.001:1, 0:0.001:1);
z = zeros(size(X,1),size(X,2));

for i = 1:length(gazedata)
z = z + gazeDuration(i) * exp(-( ((X - gazedata(i,1)).^2 ./ (2*gaussSigma^2)) + ((Y - gazedata(i,2)).^2 ./ (2*gaussSigma^2)) ) );
end

mesh(X,Y,z); % plot the heatmap
colorbar;
caxis([0,300]());
view(0,90);


print('heatmap', '-dtiff','-r300');