1. 均值滤波：即线性滤波，有:　a.邻域平均, b. 加权平均两种。
a. 邻域平均：用一个像素邻域平均值作为滤波结果，此时滤波器模板的所系数都取1，为保证输出图仍在原来的灰度值范围，在算得R后要将其除以系数总个数再进行赋值，对3*3的模板来说，在算得R后要将其除以系数9。
b. 加权平均：对同一尺寸的模板，可对不同位置的系数采用不同的数值，一般认为离对应模板中心像素近的像素应对滤波结果又较大贡献，所以接近模板中心的系数可比较大而模板边界附近的系数应比较小。(1. mean filtering: linear filtering, there are a. neighborhood averages and two B. weighted averages.
A. neighborhood average with a pixel neighborhood average value as the filtering results, the filter of the template coefficient is 1, in order to ensure the output figure is still in the original gray value range, to be divided by the total number of coefficients and then assign it in R, on the 3*3 template, to be divided by the coefficient in 9 was R.
B.: the weighted average of the same size of the template, the coefficients can be in different positions with different values, generally from the template corresponding to the center pixel near pixel filtering results and a larger contribution to deal with, so close to the center of the template coefficient is relatively large and near the boundary line template number should be relatively small.
Take the 3*3 template as an example:)