Color Image Interpolation Combined With Rough Sets Theory

Fengmei Liang,Keming Xie
DOI: https://doi.org/10.1007/978-3-540-79721-0_35
2008-01-01
Abstract:A new realtime interpolation algorithm for color image is presented. The algorithm is based on the concept of indiscernibility relation in rough sets (RS) theory. By applying the concept of upper and lower approximation based on the continuity of images, the image is first divided into homogenous area, edge pixels and isolated pixels. Then Bezier surface interpolation is further achieved using the information of classification. Besides emulation, the technology has been applied to the visual presenter with low-resolution image sensor. Results demonstrate that the new algorithm improves substantially the subjective and objective quality of the interpolated images over conventional interpolation algorithms, and meets the requirements of real time image processing. The algorithm represents an attempt to incorporate RS in image processing.
What problem does this paper attempt to address?