| 摘要 |
A whole image (51) of a low resolution is captured from an inputted image so as to accurately track a predetermined portion, an image (52) of an intermediate resolution mainly containing the head is extracted from the whole image (51), and further, after the extraction an image (53) of a high resolution mainly containing the eyes is extracted. A predetermined image different from the inputted image is changed and displayed according to the extracted image (53). |
