Foreground detection and Blob detection

Hi all ..

I'm trying to detect white objects only, but my code detect all objects

I don't know if I'm using foreground detection right or I sholud try something else !

Here is my code:

clc;
clear;
BackgroundImage = imread('frame1.jpg');
object = imread('frame2.jpg');
subplot(3,3,1);
imshow(BackgroundImage);
title('Background Frame 1');
subplot(3,3,2);
imshow(object);
title('Frame 2');
ga = rgb2gray(object);
BW = im2bw(ga);
subplot(3,3,3);
imshow(BW)
title('convert im2bw');
gb = rgb2gray(BackgroundImage);
foregroundDetector = vision.ForegroundDetector('InitialVariance',(30/255)^2);
foreground = step(foregroundDetector, gb);
subplot(3,3,4);
imshow(foreground);
title('foreground frame 1');

foreground1 = step(foregroundDetector, ga);
subplot(3,3,5);
imshow(foreground1);
title('foreground frame 2');

BlobAnalysis = vision.BlobAnalysis('MinimumBlobArea',100,'MaximumBlobArea',50000);
[area,centroid,bbox] = step(BlobAnalysis,foreground1);
Ishape = insertShape(object,'rectangle',bbox,'Color', 'green','Linewidth',6);
subplot(3,3,6);
imshow(Ishape);
title('Detect');

subplot(3,3,7);
title('Histogram');
[row , col ] = size (bbox);
for i =1 : row
 x = bbox(i,1);
 y =bbox(i,2);
 w=bbox(i,3);
 h=bbox(i,4);
TestImage = object(y :(y+h),x:(x+w), :);
r = TestImage(:,:,1);
g = TestImage(:,:,2);
b = TestImage(:,:,3);

histogram2(r,g,'DisplayStyle','tile','ShowEmptyBins','on', ...
    'XBinLimits',[0 255],'YBinLimits',[0 255]);
histogram(r,'BinMethod','integers','FaceColor','r','EdgeAlpha',0,'FaceAlpha',1)
hold on
histogram(g,'BinMethod','integers','FaceColor','g','EdgeAlpha',0,'FaceAlpha',0.7)
histogram(b,'BinMethod','integers','FaceColor','b','EdgeAlpha',0,'FaceAlpha',0.7)
xlabel('RGB value')
ylabel('Frequency')
title('Color Histogram')
xlim([0 257])

thresholding =128;
rth=graythresh(TestImage(:,:,1))*255
gth=graythresh(TestImage(:,:,2))*255
bth=graythresh(TestImage(:,:,3))*255

if ( rth*gth*bth >= thresholding )
msgbox('Alarm it is white !');
load gong.mat;
 while ( rth*gth*bth >= thresholding)%endless loop
     sound(y);
     lastTime = clock;
     while etime(clock, lastTime) < 5
        pause(0.03);
     end
    break;
 end
else
    msgbox('ok');
end
clear TestImage;
end

Here is the RESULT ! :

[Image Analyst]

You forgot to attach 'frame1.jpg'. Please do so and then we can help.

[Han]

I attach the two images

I want to detect white object after using foreground detection (Background subtraction)

Thank you

[Han]

Should I use something other than Foreground detection?

Can you please advise me?

Answer 1

[Image Analyst]

Try just comparing:

foregroundFrame2 = frame2 > backgroundFrame2;

[Han]

Thank you for your answer but sorry I didn't understand

See the images posted above.

Thank you again

[Shounak Shastri]

@Han

Image Analyst's code is correct. You would get a logical (Binary) image which highlights the white area in frame2.jpg.

Also, your images are not grayscale images, they are coloured. So you might want to add a rgb2gray before using this code as you have done in your original program.

Here's the explanation of the code:

You have, overall, 3 gray levels in your images. One is the gray which is the background in both images; the second is the white which you want to detect; and the third is the black which you don't want to detect.

The way the gray levels work is that black represents a "0" and white represents "255" for a uint8 (8-bit) image. ImageAnalyst's code creates a new image called foregroundFrame2 which contains all the pixels whose pixel values are greater than the gray in your background image marked as "1" and rest marked as "0". This separates the white block which you wanted.

[Image Analyst]

Try this:

BackgroundImage = imread('frame1.jpg');
object = imread('frame2.jpg');
subplot(3,3,1);
imshow(BackgroundImage);
title('Background Frame 1');
subplot(3,3,2);
imshow(object);
title('Frame 2');

foregroundFrame2 = object > BackgroundImage;
if ndims(foregroundFrame2) > 1
  % Convert from color to 2-D logical
  foregroundFrame2 = max(foregroundFrame2, [], 3);
end
subplot(3,3,3);
imshow(foregroundFrame2);
title('foreground Frame 2');

[Han]

Thank you so much