基于斜坡边缘模型的图像插值新方法  

New Image Interpolation Method Based on Ramp Edge Model

作　　者：张慧芯[1] 屈小波[1] 闫敬文 

机构地区：[1]厦门大学通信工程系,厦门361005 [2]广东省数字信号与图像处理重点实验室,汕头515063

出　　处：《计算机工程》2009年第10期206-208,共3页Computer Engineering

基　　金：国家自然科学基金资助项目(60472081);航空科学基金资助项目(05F07001)

摘　　要：基于斜坡边缘模型的经典插值方法把所有边缘归为强边缘,导致弱边缘过分增强而失真。针对该问题提出基于斜坡边缘模型的图像插值新方法(NIIBRED),对强弱边缘采用不同方法,考虑边缘宽度随图像放大而增大的情况,对放大图像进行修复。实验结果证明,NIIBRED使放大图像的边缘更自然且清晰,取得了更好的纹理效果。Classical interpolation method based on ramp edge model considers all the edges as strong edges, which results in weak edges＇ distortion. Aiming at this problem, a New Image Interpolation method Based on Ramp EDge model（NllBRED） is proposed, which uses different methods for strong edges and weak edges. This method considers that the edges generated in the enlarged image do not have the same width, and reconstructs the enlarged image. Experimental results show that NIIBRED can make the enlarged image＇s edges more natural and clearer, and obtain better texture effects.

关 键 词：图像插值 斜坡边缘 边缘模型 

分 类 号：TN911.73[电子电信—通信与信息系统]