Image Analysis Segmentation software engineering.pdf

DIP

1. IMAGE ANALYSIS: SEGMENTATION
INSTRUCTOR: AZKA MIR

2. IMAGE SEGMENTATION

3. WHY IS IT NECESSARY?

4. PRINCIPLE APPROACHES

5. THRESHOLDING

7. SEGMENTATION-THRESHOLDING
%load image
%declarethresholdvalues
Image=imread(“crack.jpeg”)
thresholdValue=0.3
Cracked_image= image > threshold;
Imshow(Cracked_image)

8. MULTI-LEVELTHRESHOLDING
boneThreshold= 0.4
tissueThreshold= 0.2;
bone = zeros(size(image));
tissue = zeros(size(img));
background = zeros(size(image));
bone(image>= boneThreshold) = 1;
tissue((image< boneThreshold) & (image >= tissueThreshold)) = 1;
background(image < tissueThreshold) = 1;

9. EDGE DETECTION (VERTICAL)
Img=imread(“00187.jpg”);
copied=Img;
mask=[1, 0, -1;1, 0, -1;1, 0, -1];
%Rotate image by 180 degree first flip up to down then left to right
mask=flipud(mask);
mask=fliplr(mask);
for r=2:size(Img, 1)-1
for c=2:size(I, 2)-1
N_matrix=mask.*copied(r-1:r+1, c-1:c+1);
avg=sum(N_matrix(:));
Img(r, c)=avg;
end
End;
imshow(Img)

10. SOBEL OPERATOR
maskAlongX=[-1 0 1; -2 0 2; -1 0 1];
maskAlongY=[-1 -2 -1; 0 0 0; 1 2 1];
[r,c]=size(image);
Gradient_image=zeros(r,c);
for i=2:r-1
for j=2:c-1
Gradient_along_X=sum(sum(maskAlongX.*image(i-1:i+1,j-1:j+1)));
Gradient_along_Y=sum(sum(maskAlongY.*image(i-1:i+1,j-1:j+1)));
Gradient_image(i,j)=sqrt(Gradient_along_ X^2+ Gradient_along_ Y^2);
end
end

11. PREWITT OPERATOR
maskAlongX=[-1 0 1; -1 0 1; -1 0 1];
maskAlongY=[-1 -1 -1; 0 0 0; 1 1 1];
[r,c]=size(image);
Gradient_image=zeros(r,c);
for i=2:r-1
for j=2:c-1
Gradient_along_X=sum(sum(maskAlongX.*image(i-1:i+1,j-1:j+1)));
Gradient_along_Y=sum(sum(maskAlongY.*image(i-1:i+1,j-1:j+1)));
Gradient_image(i,j)=sqrt(Gradient_along_ X^2+ Gradient_along_ Y^2);
End
End
prewittG=(prewittG/max(prewittG(:)))*255;

12. THE END