How to trace the minimum and maximum lines of the image of the moiré franges?

I have a photo of the stripes that I am posting as an example (orginal moire). How can you find the middle points of these light and dark lines by MATLAB and connect these lines with a line? I have also included an example of this image (moire with lines).

The pixel coordinates of these found lines should be stored in the matrix.

The second photo ( moire with lines) that is attached contains dark + light lines + the border between dark and light lines

Answer 1

[Catalytic]

Something like this might also work -

A=load('Image').Image;

T=findEdges(A,'top',10);
B=findEdges(A,'bottom',10);
C=findCenter(T,B);

imshow(A,[]);
hold on
for i=1:numel(T)
    h=plot(T(i).x,T(i).y,'LineWidth',2);
      plot(B(i).x,B(i).y,'LineWidth',2,'Color',h.Color);
      plot(C(i).x,C(i).y,'--','LineWidth',2,'Color',h.Color);   
end
hold off

function Line=findEdges(A,sense,n)

    outside=imerode(~bwconvhull(A>150),strel('disk',5));

    [~,S]=gradient(sgolayfilt(A,2,21,[],1));

    if strcmp(sense,'bottom'), S=-S; end

    B=(movmax(S,11,1)==S);
    B(outside)=0;
    
    B=imclose(B>0,strel('disk',4));
    B=bwareafilt(B,n);
    B=bwmorph(B,'thin');
    
    rp=regionprops(B,'PixelList');

    for i=numel(rp):-1:1

     x=rp(i).PixelList(:,1);
     y=rp(i).PixelList(:,2);

     p=polyfit(x,y,4);
     x=linspace(min(x),max(x),size(A,2));
     y=polyval(p,x);

      Line(i).x=x;
      Line(i).y=y;
      Line(i).ym=mean(y);
    end

    [~,reorder]=sort([Line.ym]);
    Line=Line(reorder);

end

function Line=findCenter(T,B)

 for i=numel(T):-1:1

   Line(i).x=(T(i).x+B(i).x)/2;
   Line(i).y=(T(i).y+B(i).y)/2;
   Line(i).ym=[];

 end

end

[Matt J]

Nice! +1

[ali alizadeh]

Wow this is wonderful!

Thank you

Answer 2

[Image Analyst]

To see published algorithms on the topic, see

9.5.3 Optical Interferometry, Moire Patterns

https://www.visionbib.com/bibliography/shapefrom395of1.html

Pick a paper and code up their algorithm.

[ali alizadeh]

Thank you for your suggestion. There are valuable articles in this link that will definitely help me.