next up previous contents
Next: Main Steps of the Up: Matching Method for Defect Previous: The Hausdorff Distance

   
Using Distance Transform to Minimize HD

Direct computation and minimization of the Hausdorff distance using its definition (2) is very time-consuming. We use the distance transform [1] of the reference shape to significantly speed up the procedure. A DT converts a binary image to another image in which pixel value is the distance from this pixel to the nearest nonzero pixel of the binary image.

The distance transform is computed off-line, and only once for each particular reference shape. We use the $5 \times 5$ neighborhood fast chamfer distance transform [1] that gives a good approximation to the exact Euclidean distance. The DT values are integers. A example of a distance transform is shown in figure 5.


  
Figure 5: A shape and its distance transform.
\begin{figure}
\begin{center}
{\epsfig{figure=/users/mitya/illustr/squash/meeti...
.../meeting4/f_letter_dt.tif.eps,width=0.3\linewidth} }
\end{center} \end{figure}



Dmitry Chetverikov
1998-11-16