Hello, why does my parallel source code turns out to be slow compared to sequential source code?

this is my sequential source code:

for n = 1:lh hj(2^j*(n-1)+1)=h(n); end

        for n = 1:lg
            gj(2^j*(n-1)+1)=g(n);
        end

        a(:,:,1,j+1) = conv2(hj,hj,a(:,:,1,j),'same');
        dx(:,:,1,j+1) = conv2(delta,gj,a(:,:,1,j),'same');
        dy(:,:,1,j+1) = conv2(gj,delta,a(:,:,1,j),'same');

        x = dx(:,:,1,j+1);
        y = dy(:,:,1,j+1);
        dj(:,:,1,j+1) = sqrt(x.^2+y.^2);

        I1 = imadjust(dj(:,:,1,j+1),stretchlim(dj(:,:,1,j+1)),[0 1]);
        figure;imshow(I1);
    end
and this is my parallel source code:
for  j = 1:J+1
      lhj = 2^j*(lh-1)+1;
      hj=zeros(1, lhj);
      hj(1:2^j:end)=h;
      a(:,:,1,j+1) = conv2(hj,hj,a(:,:,1,j),'same');
    end
   parfor j = 1:J+1
        lgj = 2^j*(lg-1)+1; 
        gj=zeros(1, lgj);
      gj(1:2^j:end)=g;
      atmp=a(:,:,1,j); %excessive optimization perhaps, eliminate one subsref() call
        dx(:,:,1,j+1) = conv2(delta,gj,atmp,'same');
        dy(:,:,1,j+1) = conv2(gj,delta,atmp,'same');
   end
    dj = sqrt(dx.^2+dy.^2);
    for j = 1:J+1
        I1 = imadjust(dj(:,:,1,j+1),stretchlim(dj(:,:,1,j+1)),[0 1]);
        figure;imshow(I1);
    end

[Jan]

Please use the "{} Code" button to post readable code.

[nuh lili]

alright. i already change it. can you read it?