computer vision,

please explain my doubt in detail, will rate the answer

why gaussian filter blur edges?

Median of a Single Set: Given this range of values 5.7.3. 4. 19,5,6,4,9 We would sort them to get: 3.4.4,5,5,6,7,9,19 Then the median of set is then 5. It would replace 19 -- which is the center value of the input set. Moving Median Filter: To make this into a full filter, we need to move the kernel along the signal Consider this signal: 1.1.1.7.2.1.1 To median filter it with a kernel of width 3 we would first find the median of the first three values (1,1) i.e. 1, then it becomes the first output value and we would move along by one pixel, repeatedly The output would then become howto achieve this sequence Note that the outlier value, 7, has been fully removed. The median of the boundary pixels is undefined -- it is up to the user to complete these as desired. 7,1,1,2,1,1,2 please explainin detail Fig. 6.1: L: Comparing the effect of 1D median filtering to mean filtering. R: 2D median filtering It does not smooth edges and true edges remain sharp. But it can also generate jagged edges by removing subtle detail, which can be unnatural. explain why generate this ? The benefit of median filtering is that it retains edges but it can cause damage to details. The Gaussian blurs edges. As seen in Fig. 6.5, an effect of Gaussian filtering is to remove definition from edges. The bilateral filter is an operation which can filter noise but does not blur edges