基于小波变换的图像插值算法

An image interpolation algorithm based on wavelet transform

提出了基于小波变换和样条函数相结合的图像插值算法.对原图像使用双三次B样条函数进行插值,对得到的插值图像做小波变换,然后对小波变换的高频子带使用Haar变换,并修改变换后的高频子带系数,进行图像重构,得到最终的插值图像.实验结果表明,用该方法对图像放大,能提高插值图像的清晰度.

In this paper a new image interpolation algorithm based on wavelet transform and spline function is proposed. We interpolate the image with bicuibc spline function at first. Secondly, we do wavelet transform to the interpolated image. Thirdly, we do Haar transform to all the high frequency subbants. Then, we modify the high frequency subbant coefficients. Lastly, we rebuilt image with modified high frequency subbant coefficients to get ultimate interpolated image. The results of the interpolated images using this algorithm are good, and the resolution of the interpolated image are improved.