hi .i'm new in matlab,plz how can i rotate all face images to be aligned vertically with same angle?

    i followed this procedure:  detecting the eyes then computing both eye centers and getting the image center,lastly get the angle between three points ,plz,could you tell me the wrong with this code as it doesn't give the appropirate angle
if true
    % code  
%%To detect the Eyes
EyeDetect = vision.CascadeObjectDetector('EyePairBig');
%Read the input Image
I = imread('a7.jpg');
A=rgb2gray(imread('a7.jpg'));
center=size(A)/2+.5;
x0 = center(1:2:end);
y0= center(2:2:end);
BB=step(EyeDetect,I);
figure,imshow(I);
rectangle('Position',BB,'LineWidth',4,'LineStyle','-','EdgeColor','b');
title('Eyes Detection');
Eyes=imcrop(I,BB);
figure,imshow(Eyes)
[a b c]=size(Eyes);
I1=Eyes(:,1:b/2);
figure,imshow(I1)
[Label,Total]=bwlabel(I1,8);
num=1;
Sdata=regionprops(Label,'centroid');
[x]= Sdata(num).Centroid;
centroids = [Sdata.Centroid];
x1= centroids(1:2:end);
y1 = centroids(2:2:end);
I2=Eyes(:,b/2:b);
[Label2,Total2]=bwlabel(I2,8);
num=1;
Sdata=regionprops(Label2,'centroid');
[y]=Sdata(num).Centroid;
centroids = [Sdata.Centroid];
x2= centroids(1:2:end);
y2 = centroids(2:2:end);
z=[x;y];
P0 = [x0, y0];
P1 = [x1, y1];
P2 = [x2, y2];
n1 = (P2 - P0) / norm(P2 -P0);  % Normalized vectors
n2 = (P1 - P0) / norm(P1 - P0);
theta = atan2(norm(det([n2; n1])), dot(n1, n2));
D = rad2deg(theta );
I=imrotate(Eyes,theta);
figure,imshow(I)

Answer 1

[Image Analyst]

Use one of the two code chunks in the FAQ: http://matlab.wikia.com/wiki/FAQ#How_can_I_process_a_sequence_of_files.3F

Then, inside the loop, you might want to call imwrite() if you want to save the rotated images to disk.