Why is Gaussian filter used in image filtering? What are its advantages compared to other filters like median filter?

Why is Gaussian filter used in image filtering? What are its advantages compared to other filters like median filter?

Answer 1

[Image Analyst]

A Gaussian filter is a linear filter. It's usually used to blur the image or to reduce noise. If you use two of them and subtract, you can use them for "unsharp masking" (edge detection). The Gaussian filter alone will blur edges and reduce contrast.

The Median filter is a non-linear filter that is most commonly used as a simple way to reduce noise in an image. It's claim to fame (over Gaussian for noise reduction) is that it removes noise while keeping edges relatively sharp.

I guess the one advantage a Gaussian filter has over a median filter is that it's faster because multiplying and adding is probably faster than sorting.

[Ender Rencuzogullari]

I am searching about filters to reduce noises for a while but I am confused little bit. My confusion still is about which filter is best to use. In order to your comments and answers in posts, I concluded that I should use wiener2 filter. However, in this post you said that median filter is most commonly used for simple tasks. On the other hand, I have used gaussian blurring for my image enhancement process and it works fine for now. However, since I will perform my code in a competition environment which is not known anything about (i.e illumination condition), I need to enhance video as much as possible for obtaining best processing result.

I can try all filters for the environment that I have built. Yet I won't be sure which one would be better for the unknown environment. All in all, how can I decide which one is better to use? Which filter would you recommend me to use for a general purpose to reduce noise and get best result according to your experience and statistics ?

Edit: before gaussian bluring, I used imadjust function to increase contrast because I skeletonize image. It gave a good result. But here is the another confusion. As i know, while gaussian bluring decreases the contrast, imadjust increases the contrast.

[Image Analyst]

Correct, that blurring reduces contrast while imadjust increases it. imadjust is not necessary - anything you do on the adjusted image could be done on the original image.

A wiener filter is used to reduce blur in the presence of noise, with blur reduction being it's main purpose. A median filter is a non-linear filter used mainly to reduce noise while not blurring edges.

I don't know your situation Ender and can't recommend a filter to either reduce noise, reduce blur, or both.

See this post: Evaluating Noise Filters by Prof Cris Luengo

and this overview paper: A modern Tour of image denoising methods by Prof Milanfar

State of the art noise reduction methods include BM3D, non-local means, K-LLD, K-SVD, and others.

Answer 2

[lourci mohamedamine]

what are the differents methods to implement the gaussian blur in image processing