difference sift matching output in matlab and opencv ?????

edit

It seems that Matlab doesn't have SIFT. Can you link to the Matlab implementation you used?

Also, is it normal that the image sizes are not the same between the two?

thank for answer images size are same.but the left picture has cropped and zoomed.it not seems that have effect on result. i think that matlab code is correct but in result of opencv is not good i need a correct opencv code about Sift image matching

enter code here

    I1=imreadbw('c:\st11.jpg') ;
    I2=imreadbw('c:\test1.jpg') ;
    I1=I1-min(I1(:)) ;
    I1=I1/max(I1(:)) ;
    I2=I2-min(I2(:)) ;
    I2=I2/max(I2(:)) ;
    fprintf('Computing frames and descriptors.\n') ;
    [frames1,descr1,gss1,dogss1] = sift( I1, 'Verbosity', 1 ) ;
    [frames2,descr2,gss2,dogss2] = sift( I2, 'Verbosity', 1 ) ;
    figure(11) ; clf ; plotss(dogss1) ; colormap gray ;
    figure(12) ; clf ; plotss(dogss2) ; colormap gray ;
    drawnow ;
    figure(2) ; clf ;
    subplot(1,2,1) ; imagesc(I1) ; colormap gray ;
    hold on ;
    h=plotsiftframe( frames1 ) ; set(h,'LineWidth',2,'Color','g') ;
    h=plotsiftframe( frames1 ) ; set(h,'LineWidth',1,'Color','k') ;

enter code here

fprintf('Computing matches.\n') ;
        descr1=uint8(512*descr1) ;
        descr2=uint8(512*descr2) ;
        tic ;
        matches=siftmatch( descr1, descr2 ) ;
        fprintf('Matched in %.3f s\n', toc) ;
        figure(3) ; clf ;
        plotmatches(I1,I2,frames1(1:2,:),frames2(1:2,:),matches) ;
        drawnow ;

this codes is image matching by sift in matlab

You should edit your question and add the C++ code you used. If you can, you should also add the individual images.

If the image of the front car for the C++ case has the same size than in your right example, you should check which criterion you used to retain the matches.

i use this code c++

http://stackoverflow.com/questions/10...

i think my code is wrong and anybody have a correct code Sift image matching by Opencv C++????