omr_techniques(matlab edition).m

clc;  
close all;  
clear;  

disp('Project in Image Processing');

%////////////
%/// MAIN ///
%////////////
  
[idx,file_name] = show_dialog();
if isempty(file_name)
    disp('Closing Program');
    return;
end
file_name = file_name{1};
switch idx
    case 1
        disp('Starting: Moire Noise Removal');
        remove_noise(file_name);
    case 2
        disp('Starting: Music StaffLine Segmentation');
        segment_lines(file_name);
    otherwise
        disp('application is exiting')
end

function remove_noise(f)
  workspace;
  format long g;
  format compact;
  directory = fileparts(which(f)); 
  filename = f;
  fullFileName = fullfile(directory, filename);
  if ~exist(fullFileName, 'file')
    fullFileName = filename;
    if ~exist(fullFileName, 'file')
      errorMessage = sprintf('Error: %s does not exist in the search path folders.', fullFileName);
      uiwait(warndlg(errorMessage));
      return;
    end
  end
  img = imread(fullFileName);
  [~, ~, numberOfColorBands] = size(img);
  if numberOfColorBands > 1
    img = rgb2gray(img);
  end

  subplot(1, 3, 1);
  imshow(img, [0, max(max(img))]);
  set(gcf, 'Name', fullFileName);
  title('Original Image', 'FontSize', 12);
  set(gcf, 'units','normalized','outerposition',[0, 0, 1, 1]); 


  %Make an image with moire noise in it.
  double_img = floor(im2double(img) * 255);
  ratio=0.1;
  double_img = padarray(double_img,[30,30],'symmetric','both');

  [m,n]=size(double_img);
  nj1 = zeros(m,n);
  nj2 = zeros(m,n);

  % N1
  for i= 1:m
      for j=1:n
          nj1(i,j)=ratio*255*(sin(1.8*i)+sin(1.8*j)+sin(i+j)+sin(2.2*i+2.2*j)+sin(1.8*i-1.8*j)+sin(i-j)+sin(2.2*i-2.2*j));
      end
  end
  % N2
  for i= 1:m
       for j=1:n
          nj2(i,j)=ratio*255*(sin(1.1*i+1.1*j)+sin(1.5*i)+sin(1.5*j+2.2*j)+sin(1.1*i-1.1*j));
       end
  end
  temp_img = double_img + nj1 + nj2;
  n_img(1:m-60,1:n-60)=temp_img(31:m-30,31:n-30);

  subplot(1, 3, 2);
  imshow(n_img/255, []);
  axis on;
  title('Image with Moire Noise Pattern', 'FontSize', 12);
  set(gcf, 'units','normalized','outerposition',[0 0 1 1]);

  temp_img=agnf(temp_img);
  output_img(1:m-60,1:n-60)=temp_img(31:m-30,31:n-30);
  subplot(1, 3, 3);
  imshow(output_img/255, []); 
  title('Image cleaned with Adaptive Gaussian Notch Filter', 'FontSize', 12);
  set(gcf, 'units','normalized','outerposition',[0, 0, 1, 1]); 
end

function segment_lines(f)
  option = questdlg('Choose music sheet element to view?', ...
	'Music Sheet Elements', ...
	'NOTES','STAVES','NOTES');
  directory = fileparts(which(f)); 
  filename = f;
  fullFileName = fullfile(directory, filename);
  if ~exist(fullFileName, 'file')
    fullFileName = filename;
    if ~exist(fullFileName, 'file')
      errorMessage = sprintf('Error: %s does not exist in the search path folders.', fullFileName);
      uiwait(warndlg(errorMessage));
      return;
    end
  end
  img = imread(fullFileName);
  [~, ~, numberOfColorBands] = size(img);
  if numberOfColorBands > 1
    img = rgb2gray(img);
  end

  img_gray = img;
  [~, pic_width] = size(img_gray);

  % binarization 
  T1 = graythresh(img_gray);
  img_gray_new = imbinarize(img_gray,T1);

  % segment intervals
  img_gray_col = sum(img_gray_new,2)/pic_width;
  T2 = graythresh(img_gray_col);
  img_gray_col_new = imbinarize(img_gray_col,T2);

  % segment
  if (isequal(option,'NOTES'))
    staff_lines = find(img_gray_col_new == 0);
  else
    staff_lines = find(img_gray_col_new == 1);
  end
  num_intervals = size(staff_lines,1)/5;
  splits = [];
  for i = 1:num_intervals-1
      splits = [splits round((staff_lines(5*i)+staff_lines(5*i+1))/2)];
  end
  %gap = round(splits(2)-splits(1))/2;
  %splits = [staff_lines(1)-gap splits];
  %splits = [splits staff_lines(num_intervals*5)+gap];

  for j = 1:length(staff_lines)
      idx = staff_lines(j);
      img_gray_new(idx,:) = 1;
  end
  fixed_img = mat2gray(img_gray_new);
  imshow(fixed_img);  
end

function [idx,file_name] = show_dialog()
  file_name = {};
  items = {'Moire Noise Removal','Music Staff Line Segmentation'};
  [idx,tf] = listdlg('PromptString',{'Select a job to execute.',...
    'Go',''},...
    'SelectionMode','single','ListSize', [250 50],'ListString',items);
  prompt = {'Enter image name:'};
  title = 'Input';
  dims = [1 35];
  if (idx == 1)
      definput = {'cameraman.tif'};
  else
      definput = {'odetojoy_min.png'};
  end
  
  if (tf ~= 0)
    file_name = inputdlg(prompt,title,dims,definput);
   end
end