Adpative image interpolation

Abstract

Simple interpolation techniques like nearest neighbor, bilinear, bicubic in the past had gained popularity due to their simplicity and low computational cost. But with the advent of high performing machines, demand for better interpolation methods at the expense of their computational complexity has arised. In this endeavor, myriads of interpolation methods have been introduced. Some of which are based on edge intensity, curvature profile of image, fuzzy logic. While others are optimized for the particular needs like resistance to outliers, performance in real time basis etc. An extensive list of interpolation methods exists in literature. We have reviewed an adaptive interpolation technique based on Newton forward dierence. This difference provides a measure of goodness for grouping of pixels around the target pixel for interpolation.