Surface-Fitting Modeling of Images: Theory and Applications
Anastase Nakassis, Abdou S. Youssef
Image processing is a very active area with many applications. Many tasks in image processing are highly complicated, and the quality of the software and hardware systems that perform those tasks has proved to be very difficult to measure. One useful approach in studying and processing complex systems is modeling. Modeling leads to simplified representation that allows for feasible analysis and targeted quality measurement. Of the many possible modeling approaches for imaging, we will choose modeling by data fitting, suitably called surface-fitting, and identify promising future directions. The applications to be considered include: (l) the design of multidimensional filters for computing derivatives, smoothing, upsampling and downsampling; (2) dynamic range adjustment of images to best meet display constraints; (3) image preprocessing techniques for better edge detection; (4) indexing of multimedia indexing for faster and more sophisticated search for audio/visual data; and (5) quality measurement of images.