Image interpolation

Abstract

In signal processing, interpolation, as one of the fundamental operations, is a significant step when up sampling images. Traditional linear interpolation methods include the nearest neighbor method, the bilinear method, and the bicubic method. Non-linear interpolation methods include wavelets method and methods based on edge detection. This paper compares the major three linear interpolation methods and two non-linear interpolation methods in details. By utilizing both the python code, the Canny method is found to provide the best performance when up sampling images.