Help with HDR code in matlab

Hello im new in matlab and im trying to figure out a code (not mine) in order to help me write my own but i keep getting this irratitng error. the code is

function [hdr, rgb, logE, points] = HDR(fileName, T)

% This function is used to generate the HDR image from the image stacks.
% Written by: Mon-Ju Wu

% Acquire the image set and decompose every image into RED, GREEN, and
% BLUE.
  if fileName(end) ~= '\'
      fileName = [fileName '\'];
  end
  fileDirect = fileName;
  FilePath = fileDirect;
  FileList = {};
  FileExt = {'jpg','JPG','jpeg','JPEG'};
  FileList = dig(FilePath,FileList,FileExt,1);

    for s = 1:size(FileList,1)
        if nargin == 1
            info = imfinfo(FileList{s});
            dc = info.DigitalCamera;
            T(s) = dc.ExposureTime;
        end
        tempImage = imread(FileList{s});
        RED(:,:,s) = tempImage(:,:,1);
        GREEN(:,:,s) = tempImage(:,:,2);
        BLUE(:,:,s) = tempImage(:,:,3);
    end

% Initial Setting.
  zMax = 255;
  zMin = 0;
  zMid = (zMax + zMin) * 0.5;
  intensityRange = 256;     % The range of intensity, which is 256.

    hdr = zeros(size(RED,1), size(RED,2), 3);

    Lamda = 100;

% Create the single weight matrix.
  for h = 1:256
      if (h-1) > zMid
          weight(h) = zMax - (h - 1);
      else
          weight(h) = (h - 1) - zMin;
      end      
  end

% Calculate the radiance.
% pixel{i}: A Nx2 matrix which records the pixels with intensity {i-1}
% points: A 256x3 matrix. The first two columns represent their locations,
% while the last column represents their intensity.
% weight: A matrix with the size of 256 which represents the weight of each
% intensity.
  for a = 1:3
      % Read in and process R, G, B separately.
      if a == 1
          colorImage = RED;
      elseif a ==2
          colorImage = GREEN;
      else
          colorImage = BLUE;
      end
      % Categorize pixels with different intensity into differeny groups.
      numPixel = 0;     % The total number of pixel that is going to be used in radiance.
      for b = 1:256
          [tempRow tempCol] = find(colorImage(:,:,round(size(colorImage,3)/2)) == b-1);
          if isempty(tempRow)
              trash = 0;
          else
              % Select a point to make sure its intensity in different
              % frame doesn't repeat.
              potentialPoints = [tempRow tempCol];
              % Randomly pick one point from "potentialPoints" and generate
              % weight matrix.
              if size(potentialPoints, 1) > 1
                  numPixel = numPixel + 1;
                  pixel{numPixel} = [tempRow tempCol];
                  pick = floor(rand * (size(potentialPoints,1) - 1)) + 1;
                  points(numPixel, :) = [potentialPoints(pick,1) potentialPoints(pick,2) (b-1)];
              end
          end
      end  

        % Generate the calculation matrices.
        A = zeros((numPixel*size(colorImage,3)) + 254 + 1, intensityRange + numPixel);
        B = zeros(size(A,1), 1);
        p = 0;
        for i = 1:numPixel
            for j = 1:size(colorImage,3)
                p = p + 1;
                A(p, colorImage(points(i,1), points(i,2), j) + 1) = weight(colorImage(points(i,1), points(i,2), j) + 1);
                A(p, intensityRange + i) = -weight(colorImage(points(i,1), points(i,2), j) + 1);
                B(p, 1) = weight(colorImage(points(i,1), points(i,2), j) + 1) * log(T(j));

            end
        end
        % Fill in the punishing terms.
        for d = (zMin + 1):(zMax - 1)
            A(numPixel*size(colorImage,3) + d, d) = Lamda * weight(d + 1);
            A(numPixel*size(colorImage,3) + d, d + 1) = -2 * Lamda * weight(d + 1);
            A(numPixel*size(colorImage,3) + d, d + 2) = Lamda * weight(d + 1);
        end
        % Add the constraint term to set the g(mid intensity) = 0
        A(end, (round(size(points,1)/2) + 1)) = 1;
        % Use SVD to solve the equation.
        X = pinv(A) * B;
        G = X(1:intensityRange);
        %lnE = X(intensityRange + 1:end);
        for q = 1:numPixel
            numerator = 0;
            denominator = 0;
            for r = 1:size(colorImage,3)
                denominator = denominator + weight(colorImage(points(q,1), points(q,2), r) + 1);
                numerator = numerator + weight(colorImage(points(q,1), points(q,2), r) + 1) * (G(colorImage(points(q,1), points(q,2), r) + 1) - log(T(r)));
            end
            lnE(q) = numerator/ denominator;
        end
        E = exp(lnE);
        logE(:,a) = lnE;
        % Normalize the Z-mid radiance.
        E = E/E(128);
        for e = 1:numPixel
            group = pixel{e};
            for f = 1:size(group, 1)
                hdr(group(f,1), group(f,2), a) = E(e);
            end
        end
    end
    imshow(hdr)
    rgb = tonemap(hdr);
    imshow(rgb)

and im getting this error:

??? Undefined function or variable "RED".
Error in ==> HDR at 36
  hdr = zeros(size(RED,1), size(RED,2), 3);

i think that RED is defined in line 25 inside the for...loop but i keep getting the error. Thanx in advance

[Shreeja Lakhlani]

Hi..I am trying to understand this code.If you can help me regarding HDR as well as this code it would be really grateful...If possible send me an e-mail regarding this.My e-mail id is [e-mail]...Thanx in advance.

Answer 1

[Walter Roberson]

RED would not be defined if dig() is unable to find the file.

I do not know of any dig() routine.

It appears to me that the routine expects a directory name to be passed in, rather than a file name, even though it names the parameter fileName

[Thrasyvoulos]

() is another function i didnt post it.anyway thnx for the answer i will look into it and try to solve the issue. here is dig():

function [FileList] = dig(FilePath,FileList,FileExt,verbose)
dirList = dir(FilePath);
ridx = strcmp({dirList.name},'.') | strcmp({dirList.name},'..');
dirList(ridx) = [];
if size(dirList,2) ~= 0
    for listing = 1:size(dirList,1)
        current_Path = [FilePath dirList(listing).name];
        typed_path = regexprep(current_Path,'\','\\\');
        if verbose
            fprintf(['Looking at:' typed_path '\n']);
        end
        if dirList(listing).isdir
            FileList = dig([current_Path '\'],FileList,FileExt,verbose);
        else
            FileExt;
            dirList(listing).name(end-2:end);
            if any(strcmp(FileExt,dirList(listing).name(end-2:end)))
                FileList{end+1,1} = current_Path;
            end
        end
    end
end