Attempt to reference field of non-structure array.

whenever i run the code with parameter(3,2,2*19,4,5)..it gives error attempt to reference field of non structure array..please help to resolve error

function [x,y]=Untitled2( x, y, IntegralImages, w,h)
% [x,y]=Untitled2( x, y, IntegralImages, w,h)
%

% Calculate the mean 
InverseArea = 1 / (w*h);
mean =  GetSumRect(IntegralImages.ii,x,y,w,h)*InverseArea;
% Use the mean and squared integral image to calculate the grey-level
% Varianceiance, of every search window
Variance = GetSumRect(IntegralImages.ii2,x,y,w,h)*InverseArea - (mean.^2);

Answer 1

[Azzi Abdelmalek]

What is IntegralImages.ii? this causes the error

[divya]

i Want to calculate the mean and variance of image using integral images ..it is not giving me correct answers..can u tell me the error?

function [x,y]=Untitled2( x, y, w,h)
% [x,y]=Untitled2( x, y, IntegralImages, w,h)
function IntegralImages = GetIntergralImages(Picture)
IntegralImages=GetIntergralImages('RunAppDownload.jpg');
Picture=double(Picture);

% Resize the image to decrease the processing-time
%if(Options.Resize)
%   if (size(Picture,2) > size(Picture,1)),
%      Ratio = size(Picture,2) / 384;
 % else
  %    Ratio = size(Picture,1) / 384;
  %end
  %Picture = imresize(Picture, [size(Picture,1) size(Picture,2) ]/ Ratio);

%else
%   Ratio=1;
%end
% Convert the picture to greyscale (this line is the same as rgb2gray, see help)
if(size(Picture,3)>1),
  Picture=0.2989*Picture(:,:,1) + 0.5870*Picture(:,:,2)+ 0.1140*Picture(:,:,3);
end
% Make the integral image for fast region sum look up
IntegralImages.ii=cumsum(cumsum(Picture,1),2);
IntegralImages.ii=padarray(IntegralImages.ii,[1 1], 0, 'pre');
% Make integral image to calculate fast a local standard deviation of the
% pixel data
IntegralImages.ii2=cumsum(cumsum(Picture.^2,1),2);
IntegralImages.ii2=padarray(IntegralImages.ii2,[1 1], 0, 'pre');
% Store other data
IntegralImages.width = size(Picture,2);
IntegralImages.height = size(Picture,1);
%IntegralImages.Ratio=Ratio;

% Calculate the mean 
InverseArea = 1 / (w*h);
mean =  GetSumRect(IntegralImages.ii,x,y,w,h)*InverseArea;
% Use the mean and squared integral image to calculate the grey-level
% Varianceiance, of every search window
Variance = GetSumRect(IntegralImages.ii2,x,y,w,h)*InverseArea - (mean.^2);

[Jan]

@divya: No, it is your turn to tell us the error. How could we guess, what the expected "correct" result is?