基于小波包变换的尺度自适应阈值图像降噪

针对传统小波变换易引起图像边缘模糊的不足,研究了基于小波包变换的尺度自适应阈值图像降噪。利用小波包变换可以同时对图像的高频和低频部分进行进一步的细分,因此可以更好地保留原图像信息的优势,更进一步,克服通用阈值的缺陷,软阈值函数的构造充分考虑了不同尺度层次上的系数的不同特点,产生尺度自适应阈值。通过对加噪图像的实验可以看出,与传统方法相比,本文方法不仅降噪效果有很大的改进,而且有很好的视觉效果,峰值信噪比也有较大幅度的提高。

Scale Adaptive Threshold Image De-noising Based on Wavelet Packet

Traditional wavelet transform will make the edge of image blurry. Aiming at this kind of disadvantage, scale adaptive threshold image de-nolsing which is based on wavelet packet transform is researched. Low frequency part and high frequency part can be decomposed at the same time by wavelet packet transform, so it can retain original image information better. This advantage is utilized and furthermore, the shortcoming of general threshold is overcome. The construction of soft threshold considers the different characteristic of the coefficients in different scales. Therefore, scale adaptive threshold is produced. Via the experiment on noisy image, we can see that compared with traditional methods, our method not only greatly improve the effect of de-noising, but also gain good visual effect. The peak signal to noise ratio is also enhanced.