摘要
图像插值是一个从低分辨率图像重建出高分辨率图像的技术。保持图像结构边缘的锐利和更多的图像细节是图像插值技术的具有挑战性的任务。提出一种基于中值的图像插值方法,该方法利用中值滤波器分步得到高分辨图像的灰度值。实验仿真表明所提出的方法可以避免传统线性插值出现的过模糊,振铃,锯齿等假象,同时该算法的计算复杂度较低。
Image interpolation is a technique to reconstruct a high-resolution （HR） image from the associated low-resolution （LR） input image. To preserve the sharpness of the edges structure of image and more image details are challenging task to the image interpolation technique. In this paper, we propose a new median-based image interpolation algorithm. It uses the median filter to get gray values of the HR image step by step. Simulation experiment shows that the proposed algorithm can prevent the artifacts of over blurring, ringing, jags, etc. the traditional linear interpolation has, and the computation complexity of the algorithm is lower as well.
出处
《计算机应用与软件》
CSCD
北大核心
2014年第7期229-232,共4页
Computer Applications and Software
关键词
图像插值
非线性
中值滤波
灰度值
Image interpolation
Nonlinear
Median filter
Grey value
