| 摘要 |
<p>A feature-point verification unit calculates a first verification score between a first organism pattern and a second organism pattern, on the basis of a first feature point data that indicates a first feature-point group of a first ridge pattern contained in a first image of the first organism pattern, and a second feature point data that indicates a second feature-point group of a second ridge pattern contained in a second image of the second organism pattern, extracts, from the first feature-point group and the second feature-point group, a group of feature points corresponding to each other as a corresponding feature-point group, and generates a corresponding feature-point list. An image nonlinear-conversion unit executes a first-image nonlinear conversion that makes the first image become more similar to the second image, on the basis of the corresponding feature-point list. The feature-point verification unit calculates a second verification score between the first organism pattern and the second organism pattern, on the basis of the first image and the second image after the first-image nonlinear conversion was executed. In such a way, high verification precision can be achieved in finger-print verification or palm-print verification that uses low-quality images.</p> |
