There are so an image, you can see that there is a lot of noise on the image:
High-frequency signal as if the ground towering peaks:
It looks very conspicuous.
One way is to smooth this mountain, the mountain dig out some earth, fill to go around the mountain. Mathematically words, is the average height of the peaks surrounding it.
Smoothing obtained:
4.2 computing
Convolution can help achieve this smoothing algorithm.
There is noise in the picture, you can put it into a matrix:
Then the following average matrix (described, the picture processing is actually in the normal matrix, where for simplicity, to use the matrix arithmetic mean) to smooth the image:
Remember just said algorithm, the high-frequency signal with an average value around what you can smooth the peaks.
For example, after the smoothing a1,1a1,1 points I, in the matrix in the vicinity of the withdrawal point dots a1,1a1,1 matrix F, g, and subjected to convolutional calculation, reloading back
To note that, in order to use convolution, g and f although the same dimensions, but the subscript a bit different: