摘要 |
PROBLEM TO BE SOLVED: To automatically generate a partial template matching partial data of each learning pattern, even if the position of the partial data on the learning pattern is unknown. SOLUTION: This device has a means 2 which extracts the partial data matching the learning pattern 1 the most from the learning pattern 1 with shifting the position of the partial data in each the learning pattern 1; and a means 3 for deciding a partial template 4 by use of the extracted partial data. The device automatically generates the partial template 4 in each category from the learning pattern 1, where the partial data is unknown. Because the partial data matching the most is extracted through shifting of the position in each the learning pattern 1, the partial template 4 matching the partial data of each the learning pattern 1 can be generated automatically, even if the position of the partial data in the learning pattern 1 is unknown. By collating the partial template 4 thus generated with an input pattern, the input pattern where the part of data is lost can be categorized.
|