边缘保真去噪时窗分解新方法

在常规的图像处理中,为了减小噪音、提高信噪比,一般的处理方法中都含有平均效应,会造成信息的边界发生严重的畸变,降低分辨率。在围绕时窗分解方面提出了矩形或长方体分解方法与前人提出的多边形或多面体分解方法不同,实例计算结果对比分析,证明效果同样显著,并且矩形或长方体分解方法不容易出错,更为实用。

The new method of decomposition of time window of edge-preserving smoothing

In the conventional image processing ,in order to reduce noise and improve signal to noise ratio ,the general processing method contains average effect,and the boundary information will distort severely and the resolution will reduce. Luo Yi etc put forward the edge-preserving smoothing algorithm ,which subtly ease the contradiction between signal to noise ratio and resolution,it obtain the mean value corresponding minimum standard deviation to reduce noise and achieve effective fidelity of step sharp edge.This method Significant effect at the edge detection(such as faults,fractures).This paper put forward rectangular or cuboid decomposition method during decomposition of time window,which is different from the method of previous proposed polygon or multilateral body decomposition ,the analysis of examples of comparative of results prove that the effect is the same remarkable,and the rectangular or cuboid decomposition method does not make mistakes and more practical.