Gaussian blur

• A filtering technique that uses a Gaussian function to create a smoothing effect on an image. The blur reduces image noise and reduces detail by averaging pixel values with their neighbors.

- it suppresses high frequency noise, keeps larger scale structure, makes gradients reflect real boundaries.

- When we apply Gaussian Blur, we create a kernel (a small matrix). The kernel is then convolved with the image, which means each pixel’s new value becomes a weighted average of itself and its neighboring pixels, with weights determined by the Gaussian function.

- When we apply Gaussian Blur, the pixel intensities get averaged with surrounding pixels, with closer pixels having more influence than farther ones.

For example, if we have a sharp edge in an image:

 

Before Gaussian Blur:
10 10 10 | 200 200 200
10 10 10 | 200 200 200
10 10 10 | 200 200 200

 

After Gaussian Blur:

10 20 50 | 150 190 200
10 20 50 | 150 190 200
10 20 50 | 150 190 200

 

The sharp edge becomes a gradual transition. Pixels near the edge get averaged, creating a smoother gradient.