I am trying to compress an image through DCT compression, but i have some problems on the idea of collecting zeroz and non zeros values for each 8*8 sub_blocks in order to get the probabilty for the huffman conding, can anybody help me please?

....... ........ for i=1:128 for j = 1:128

        t=q_ca(i,j);
        t=cell2mat(t);
        b=ZigZagScan(t);

        vect=b

        count0=0;

        l = 1;
       p=1;
        for o= 1:64
            if vect(o)==0

                count0=count0 + 1;

            elseif vect(o) ~=0
                    noZ(p)=vect(o)
                    p=p+1;
                    if count0 ~= 0
                        Z(l)= count0  
                        l=l+1;
                        count0=0; 

                    end    
            end
 end
        Z=0;
   end
end
......................