An Adaptive Image Interpolation Method Focusing on Edge Information

摘要

In this paper, an adaptive interpolation method of digital image focusing on edge information is studied. The target is multi-level zoom ensuring image quality and keeping edge information. And this method is composed of three key models. (1) Accurate edge detecting model. Edge detecting model is the basic model to describe image features and the classic Canny edge detector is applied to get accurate image edge information. (2) Adaptive interpolation model. Kriging interpolation algorithm is used to interpolate zoom image using gray information and spatial distribution information of image. (3) adaptive image interpolation. Using Canny edge detector, original image is divided into edge and non edge region. And in image zoom process, the weighted neighborhood interpolation is applied to produce non edge pixels. Kriging interpolation is applied to produce edge pixels to Keep edge characteristics. Experiment is carried out to test the effectiveness and characteristics of this approach. Experimental results approve that it can realize multi-level zoom and keep edge information while image zoom out. Experimental results show the validity and practicability of the research method and it has good application potential.