论文部分内容阅读
文中提出了一种新型的递归中值滤波器，将滤波算法转化为一种优化处理，兼顾了滤波处理的光滑连续性及抑制噪声的累积特性，可有效地消除脉冲型干扰的影响，并对该算法进行了分析，为消除加性高斯噪声，提出了一种小波线性滤波器，它仅仅处理边缘信息，克服了边缘模糊效应，小波的去噪逆向重构的处理方法，对边缘为阶跃型的层析图像非常实用。
In this paper, a new type of recursive median filter is proposed, which transforms the filtering algorithm into an optimization process, taking into account the smooth continuity of the filtering process and the cumulative characteristics of suppressing noise, which can effectively eliminate the influence of impulsive interference. The algorithm is analyzed. In order to eliminate the additive Gaussian noise, a wavelet linear filter is proposed. It only deals with the edge information, overcomes the edge blurring effect and the denoising inverse reconstruction of the wavelet. Type of tomography is very useful.