Limited-Region Refinement

Robert P. Munafo, 1993 Dec 23.

"Limited-Region Refinement" refers to various user-interface techniques through which a user can selectively refine regions of an image while the image is still being drawn.

In a typical implementation, a user could select a small rectangle of the image (with a mouse). The program would stop its normal imaging, image the contents of the rectangle, and then continue with its normal imaging. Depending on the imaging algorithm being used, "image the contents of the rectangle" could mean lots of different things. For example, a program that uses Successive Refinement (q.v.) would scan the rectangle with successive refinement, thereby doubling the resolution of everything inside the rectangle. A program using Automatic Dwell Limit might iterate the points inside the rectangle for 2N iterations, where N is the current number of iterations that have been done on the whole image.