| 摘要 |
When a comparison is made between each partial image in frame images output in a frame order from a solid-state image capturing element and plural types of face patterns 25, 26, which are prepared in advance, according to pattern matching to extract an area of a person face image existing in each frame image as a face candidate, the pattern matching is performed with limiting face patterns used in extracting a face candidate from a current frame image to face patterns, of the plural types of face patterns, used in extracting a face candidate from a preceding frame image (step S 29 ).
|
