摘要
图像放大技术的主要难点在于如何使放大后的图像尽可能保持原始图像的清晰度。传统的内插法存在一定的缺陷。提出了基于小波变换的ENO插值图像放大方法,对插值后的图像做小波变换,对于高频系数进行修正,经过重构保持放大图像的清晰度。实验结果表明,用该方法与传统的图像放大方法相比,获得的放大图像的纹理特征和图像的边缘得到明显增强,并能提高插值图像的清晰度。
The main difficulty of image magnification technique lies in how to ensure the definition of the amplificatory image because traditional interpolation algorithm has shortcomings. A method of image magnification with ENO interpolation which based on wavelet transform is presented, to enhance the coefficient of the high frequency subband. Coefficient recon- struct can keep the definition of the amplificatory image. The experimental results show that the method has better performance than the other methods, for retaining image texture and intensifying image edge.
出处
《现代电子技术》
2007年第24期135-136,142,共3页
Modern Electronics Technique
关键词
ENO插值
小波变换
图像放大
分形插值
ENO interpolation
wavelet transformation
image magnification
fractal interpolation