2D CFAR algorithm.

I aim is to contruct a training region, gaurd region around a cell under test in the image. The mean and standard deviation of the training region are to be compared to the cell under test.

B= training band

G= guard band

P= cell under test

k=threshold.

for i = Tr-Gr:Tr-Gr
     for j =Td-Gd: Td-Gd
        
         for p = i-14 : i-14
             for q = j-12 : j-12
               CUT=I(p,q);
               target=I(i,j);
               mt=mean(target(:));
               stdt=std(target(:));
               test=(abs(CUT-mt)/stdt);
         if (test < K)
            CUT = 0;
       else
           CUT = 1;
         end
             end 
                          
         end
     end
end

In the output I'm expecting a binary image with detected ship, but im unable to obatin the output.

[Mehmed Saad]

2D CFAR but it is on Data and not on image

Also in your code p and q are negative due to wrong use of Tr and Td

[Sanika Vivek Thakar]

I had tried to use that documentation but couldn't figure how to do it for an image. thank you for your suggestion.

[Image Analyst]

Can you attach the 'imagefile' image?

Answer 1

[Image Analyst]

Sorry, it's been so long you probably already have this, but anyway for comparison purposes, here is my version. Change the image file name or use uigetfile() to select it.

Try this:

% Demo to find ships.  By Image Analyst, May 5, 2020.
clc;    % Clear the command window.
close all;  % Close all figures (except those of imtool.)
clearvars;
workspace;  % Make sure the workspace panel is showing.
format long g;
format compact;
fontSize = 16;
fprintf('Beginning to run %s.m ...\n', mfilename);

% Read in image.
folder = pwd;
baseFileName = 'ships.jpeg';
fullFileName = fullfile(folder, baseFileName);
% Check if file exists.
if ~exist(fullFileName, 'file')
	% The file doesn't exist -- didn't find it there in that folder.
	% Check the entire search path (other folders) for the file by stripping off the folder.
	fullFileNameOnSearchPath = baseFileName; % No path this time.
	if ~exist(fullFileNameOnSearchPath, 'file')
		% Still didn't find it.  Alert user.
		errorMessage = sprintf('Error: %s does not exist in the search path folders.', fullFileName);
		uiwait(warndlg(errorMessage));
		return;
	end
end
rgbImage = imread(fullFileName);
[rows, columns, numberOfColorChannels] = size(rgbImage);
% Display the original image full size.
imshow(rgbImage, []);
axis('on', 'image');
caption = sprintf('Original Color Image : "%s"', baseFileName);
title(caption, 'FontSize', fontSize, 'Interpreter', 'None');
drawnow;
hp = impixelinfo(); % Set up status line to see values when you mouse over the image.

% Set up figure properties:
% Enlarge figure to full screen.
hFig = gcf;
hFig.Units = 'Normalized';
hFig.WindowState = 'maximized';
% Get rid of tool bar and pulldown menus that are along top of figure.
% set(gcf, 'Toolbar', 'none', 'Menu', 'none');
% Give a name to the title bar.
hFig.Name = 'Demo by Image Analyst';
hFig.NumberTitle = 'Off'

% Let user draw the guard region.
uiwait(msgbox('Draw the guard region'));
guardRect = drawrectangle('Color', 'green');

% Let user draw the guard region.
uiwait(msgbox('Draw the P region'));
pRect = drawrectangle('Color', 'yellow');

% Save the position because when we redisplay the image, the rects will vanish otherwise.
gRect = guardRect.Position;
pRect = pRect.Position;

% Display the original image.
subplot(4, 2, 1);
imshow(rgbImage, []);
axis('on', 'image');
caption = sprintf('Original Color Image\n"%s"', baseFileName);
title(caption, 'FontSize', fontSize, 'Interpreter', 'None');
drawnow;
hp = impixelinfo(); % Set up status line to see values when you mouse over the image.
% Draw the rectangles again.
rectangle('Position', gRect, 'EdgeColor', 'g', 'LineWidth', 2);
rectangle('Position', pRect, 'EdgeColor', 'y', 'LineWidth', 2);

grayImage = rgb2gray(rgbImage);
% Display the gray scale image.
subplot(4, 2, 2);
imshow(grayImage, []);
axis('on', 'image');
caption = sprintf('Gray Scale Image');
title(caption, 'FontSize', fontSize, 'Interpreter', 'None');
drawnow;
hp = impixelinfo(); % Set up status line to see values when you mouse over the image.
% Draw the rectangles again.
rectangle('Position', gRect, 'EdgeColor', 'g', 'LineWidth', 2);
rectangle('Position', pRect, 'EdgeColor', 'y', 'LineWidth', 2);

% Make guard mask
gRect = int32(gRect);
guardMask = true(rows, columns);
colg1 = gRect(1);
colg2 = gRect(1) + gRect(3);
rowg1 = gRect(2);
rowg2 = gRect(2) + gRect(4);
guardMask(rowg1:rowg2, colg1:colg2) = false;
% Get the mean and standard deviation in the mask region.
gMean = mean(grayImage(guardMask))
gStd = std(double(grayImage(guardMask)))
% Display the mask image.
subplot(4, 2, 3);
imshow(guardMask, []);
axis('on', 'image');
caption = sprintf('Guard Mask Image, Mean = %.2f, StDev = %.2f', gMean, gStd);
title(caption, 'FontSize', fontSize, 'Interpreter', 'None');
drawnow;

% Make guard mask
pRect = int32(pRect);
pMask = false(rows, columns);
colp1 = pRect(1);
colp2 = pRect(1) + pRect(3);
rowp1 = pRect(2);
rowp2 = pRect(2) + pRect(4);
pMask(rowp1:rowp2, colp1:colp2) = true;
% Get the mean and standard deviation in the mask region.
pMean = mean(grayImage(pMask))
pStd = std(double(grayImage(pMask)))
% Display the mask image.
subplot(4, 2, 4);
imshow(pMask, []);
axis('on', 'image');
caption = sprintf('P Mask Image, Mean = %.2f, StDev = %.2f', pMean, pStd);
title(caption, 'FontSize', fontSize, 'Interpreter', 'None');
drawnow;

% Extract P image
P = double(grayImage(rowp1:rowp2, colp1:colp2));
% Display the P image.
subplot(4, 2, 5);
imshow(P, []);
impixelinfo;
axis('on', 'image');
caption = sprintf('P Image, Mean = %.2f, StDev = %.2f', pMean, pStd);
title(caption, 'FontSize', fontSize, 'Interpreter', 'None');
drawnow;

% Compute ratio
ratio = abs(P - gMean) ./ gStd;
% Display the P image.
subplot(4, 2, 6);
imshow(ratio, []);
impixelinfo;
axis('on', 'image');
caption = sprintf('Ratio Image');
title(caption, 'FontSize', fontSize, 'Interpreter', 'None');
drawnow;

% Threshold the image.
kThreshold = 1.01;
shipMask = ratio > kThreshold;
% Display the shipMask image.
subplot(4, 2, 7);
imshow(shipMask, []);
impixelinfo;
axis('on', 'image');
caption = sprintf('Ship Mask Image with Threshold = %.2f', kThreshold);
title(caption, 'FontSize', fontSize, 'Interpreter', 'None');
drawnow;

fprintf('Done running %s.m\nThank you Image Analyst!', mfilename);

[Sanika Vivek Thakar]

Thank you so much, this really helps. Is there a way I can automate the drawing of boxes?

[Image Analyst]

Sure, but like how??? Do you just want to compute the edges based on some fraction in from the edges of the image? Like row1 = round(0.25 * rows) or whatever?

[Sanika Vivek Thakar]

yes, by stating the rows and columns of the gurad and p region.

[Image Analyst]

Did you see the lines:

colp1 = pRect(1);
colp2 = pRect(1) + pRect(3);
rowp1 = pRect(2);
rowp2 = pRect(2) + pRect(4);

Well, just replace the right hand side of those equations with whatever numbers you want,

colp1 = 300; % or whatever you want.
colp2 = 400;
rowp1 = 350;
rowp2 = 450;

and delete the call to drawrectangle().

[Sanika Vivek Thakar]

Sure, thank you much for your valuable input to my query.