一种基于联合插值的图像旋转算法  

The Image Rotation Methods Based on Combination Interpolation

作　　者：陈向奎[1] 李永亮[1] 林晓[1] 

机构地区：[1]洛阳师范学院信息技术学院,河南洛阳471022

出　　处：《河南大学学报（自然科学版）》2015年第1期90-94,101,共6页Journal of Henan University:Natural Science

基　　金：国家自然科学基金项目(U1304616)

摘　　要：针对现有的图像插值方法在图像旋转中容易出现"锯齿"形边缘和模糊的现象,采用新的图像表示模型,提出平面插值、球面插值、斜平面插值、双三次插值相结合的联合图像插值方法,根据待插入点周围像素点空间位置的不同,分别使用不同的插值方法,从而避免了传统的插值方法用同一个模型逼近所有像素存在的不足.示例分析和实验结果表明了联合插值方法的有效性.The existing image interpolation methods have some shortcomings,i.e.,the image blur or saw-tooth shaped edges in rotated image.In the paper,the image representation model has been modified,and the image combination interpolation methods are given including the plane interpolation,spherical interpolation,slope interpolation and bicubic interpolation in order to avord the shortcomings of the traditional interpolation methods in which all the pixels are approached with the same model.This paper selected different interpolation method according to different pixel space positions around the wanted insert point.Through the example analysis and experimental results,the given methods of image interpolation are shown effectiveness.

关 键 词：图像插值 平面插值 球面插值 斜平面插值 双立方插值 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]