| 摘要 |
<p>Embodiments of the present invention seek to improve motion detection accuracy. A moving region discriminator of a motion detector circuit utilizes the field differential value and three frame differential values a through c to detect a moving region while also taking scene change, device mode information and line differential values into account. Frame differential values b and c are calculated utilizing the pixel positions on the lines above and below the line positioned with pixels for the calculated frame differential a. An isolation region eliminator detects a moving region enclosed by still regions, and converts that moving region into a still region. A moving region expander converts into moving regions, a specified number of pixels of still regions adjoining regions determined to be moving regions. <IMAGE></p> |
