| Popis: |
Face recognition often fails when the target face is hidden or angled. Therefore, entry/exit management systems using face recognition force users to perform some operation or a temporary stop for recognition. In work environments where users frequently enter and leave the room, it is very inconvenient to force such operations and actions. In this research, we develop personal identification methods using facial features and closing features as methods without user operation. We propose several methods that differ in the integration of facial and clothing recognition results. The evaluation results demonstrate the effectiveness of the proposed methods. |
