| 摘要 |
A plurality of linear images are cut, the cut linear images are scanned successively on other images to detect positions where the degree of similarity is high, and these detected coordinate values are partially combined, and when the image is collated, combinations which does not satisfy the relative positional relation corresponding to these linear images are excluded from the processing target with reference to the relation between coordinate values of the detected positions. The image collation device thus enables a fingerprint collation device to obtain collation result consistently within a short time.
|
