Code covered by the BSD License
-
[L]=lineint(I,C,r,n,part)
function to calculate the normalised line integral around a circular contour
-
[O]=drawcircle(I,C,r,n)
function to generate the pixels on the boundary of a regular polygon of n sides
-
[b,r,blur]=partiald(I,C,rmin,...
function to find the partial derivative
-
[ci,cp,out]=thresh(I,rmin,rma...
function to search for the centre coordinates of the pupil and the iris
-
[cp]=search(im,rmin,rmax,x,y,...
function to detect the pupil boundary
-
getEigenEye.m
% Purpose : Detect Pupil from given image
-
getEye(img,out)
% Purpose : find position of Pupil within an Image
-
getEyes(left,right)
% Purpose : Detection of both Pupils from two detail images
-
getFeatures.m
% Purpose : Find boxes of interest for Eyes, Mouth and Nose
-
getMouth(pic)
% Purpose : find position of Mouth within an Image
-
getNose(pic)
% Purpose : find position of Mouth within an Image
-
getPoints.m
% Purpose : Detection distinct points for Mouth, Nostrils, Eyes
-
getResolution.m
% Purpose : Finds the highest possible resolution of an video input device.
-
eyetracking.m
-
View all files
from
Fast Eyetracking
by Peter Aldrian
Extract feature points from faces to track eye's
movement
|
| getEigenEye.m |
function[out]=getEigenEye(img)
%
% Purpose : Detect Pupil from given image
%
% Uses (syntax) :
% [out]=getEigenEye(img)
%
% Input Parameters :
% img := RGB-Image (m-by-n-by-3 matrix); bounding box of Eye
%
% Return Parameters :
% out := structure
% out.si := size of img
% out.o := img with marked pupils (m-by-n-by-3 matrix)
% out.pic := EigenPupil (m-by-n-by-3 matrix)
%
% Description and algorithms:
% Using algorithm of Anirudh S.K. (thresh.m) to detect Pupil (which takes
% rather long).
%
% Author : Peter Aldrian, Uwe Meier, Andre Pura
% Date : August 12, 2009
% Version : 1.0
% -------------------------------------------------------------------------
% (c) 2009, Meier, University of Leoben, Leoben, Austria
% email: aldrian.peter@gmail.com uwemei@gmail.com andre.pura@gmail.com
% -------------------------------------------------------------------------
s = size(img);
out.si=s;
[ci,cp,o]=thresh(img,s(1) * 0.1,s(1)*0.3);
ci=round(ci);
out.o = o;
out.pic=(img(ci(1)-ci(3):ci(1)+ci(3),ci(2)-ci(3):ci(2)+ci(3),:));
end
|
|
Contact us at files@mathworks.com