You are viewing the course site for a past offering of this course. The current offering may be found here.
Lecture 24: Image Processing (30)
sethzhao506

Some common blurring methods include averaging, Gaussian Blur, median Blur and Bilateral Filtering. OpenCV has documentation where you could try it yourself: https://docs.opencv.org/master/d4/d13/tutorial_py_filtering.html.

wcyjames

Gaussian Blur can be seen as a low-pass filter that is used to remove "Gaussian" or random noise from the image. It can still be understood as taking the average value of its surrounding pixels.

Leon-Shao

@wcyjames I don't quite know how taking the average value of surrounding pixels can remove random noise from the image, I guess taking the average reduces the error?

ja5087

I'd say the part about taking the average is better described as a box filter, and for that if you think about it, the variance is reduced by a factor of n given independently distributed zero-mean noise, and with large enough samples the central limit theorem will apply. The gaussian filter is more like a 2D grid of Gaussian PDF samples and I found a derivation here

alexkassil

Box filters will lead to nasty artifacts that guassian blur doesn't have

You must be enrolled in the course to comment