A Survey on Non Adaptive Image Interpolation Algorithms Based on Quantitative Measures

Image interpolation is a procedure used in expanding and contracting digital images. Image interpolation is referred by various terminologies such as image scaling, image resizing, image resampling, digital zooming, image magnification or enhancement. Image interpolation is widely used in many consumer application devices such as charge-coupled device (CCD) camera, closed circuit television (CCTV), mobile phones, high definition players and endoscopes. There are two kinds of image interpolation algorithms as non-adaptive and adaptive interpolations. The non-adaptive method uses a fixed methodology to handle all source pixels throughout the interpolation operation. But the adaptive technique uses different techniques to handle source pixels during the entire interpolation process. The advantage of adaptive technique is better interpolated image, but it requires high computational cost. So to reduce the computational cost of image interpolation, nonadaptive interpolation algorithms are preferred. This survey analysis the characteristics of different non-adaptive interpolation techniques based on their peak signal to noise ratio (PSNR) of interpolated image and their computational complexity. This work also explores the applicability of these techniques in real time applications. Based on the evaluation of different interpolation techniques it can be suggested that the first order polynomial convolution interpolation (FOPCI) is suitable for real time applications due to its better PSNR and its low computational complexity, and the performance of FOPCI can also be improved by using appropriate filters.