摘要 |
In this image processing device (3), temporally continuous first through n-th frame images (n being an integer two or greater) based on video images are sequentially received, there being provided: a rough extraction determination unit (130) for sequentially determining first through n-th rough extraction regions in the first through n-th frame images by rough-extracting a lung field region and carrying out a rough extraction process using a template; and a precise extraction determination unit (140) for sequentially determining a first through n-th precise extraction region in the first through n-th frame images by carrying out a precise extraction process of the lung field region from the first through n-th rough extraction regions. The template used in the rough extraction process of an i-th frame image (any integer where 2 <= i <= n) is dynamically modified on the basis of the rough extraction process result or precise extraction process result of the (i-1)-th frame image. This efficiently enhances the precision of extraction when a target region is being extracted from video images. |