Feature extraction by shape matching - Feature Extraction and Image Processing - page 204

the Cartesian definition given in Equation 5.74. According to this equation the displacement
vector is given by
γ (1, 0) = ω (θ ) - b (5.84)
The matrix T stores the co-ordinates of γ (1, 0). This matrix is expanded to accommodate
all the computed entries.
Code 5.11 shows the implementation of the gathering process of the GHT. In this case
we use the Cartesian definition in Equation 5.74. The co-ordinates of points given by

%Generalised Hough Transform

function GHT(inputimage,RTable)

%image size
[rows,columns]=size(inputimage);

%table size
[rowsT,h,columnsT]=size(RTable);
D=pi/rowsT;

%edges
[M,Ang]=Edges(inputimage);
M=MaxSupr(M,Ang);

%accumulator
acc=zeros(rows,columns);

%for each edge point
for x=1:columns
for y=1:rows
if(M(y,x)~=0)

phi=Ang(y,x);
i=round((phi+(pi/2))/D);
if(i==0) i=1; end;

for j=1:columnsT
if(RTable(i,1,j)==0 & RTable(i,2,j)==0)
j=columnsT; %no more entries
else
a0=x-RTable(i,1,j); b0=y-RTable (1, 2, j);
if(a0>0 & a0<columns & b0>0 & b0<rows)
acc(b0,a0)=acc(b0,a0)+1;
end
end
end
end %if
end %y
end %x

Code 5.11 Implementing the GHT