| 摘要 |
An image alignment method and system for aligning two digital images which were intended to be scanned from the same portion of medium, but which may be skewed somewhat. A portion of the first digital image is selected as the target region and a portion of the second digital image is selected as a variable or movable region. Data from the two regions can be subjected to a mathematical or statistical algorithm which determines how well the data from the regions match or correlate. Then, a new region can be selected to serve as the variable region from the second digital image and the correlation algorithm can be conducted again. This process can be repeated on a number of regions from the second digital image. Regions having the highest mathematical correlation can then be used to determine offsets which indicate the amount of skew or misalignment between the two digital images.
|
