摘要 |
<p>A pattern matching particularly useful for recognizing an input string of spoken words is described. The method includes the step of determining the distance d(m, j) between each feature vector b? of a reference pattern and a feature vector am of an input pattern at each point (m, j) in the i-th block defined in the form of a parallelogram which has the IL frame width of an input pattern and is inclined with respect to the reference pattern time axis. DP matching calculations are then conducted on the basis of the distance determined. Then there follows the step of setting the boundary value of an (i+l)-th slant block at values at such a slant line contacting with an i-th block as is obtained by the DP matching thereby to effect the DP matching calculations.</p> |