I think Gaussian blur is a really cool technique that can be used to preprocess images. However I’m wondering what is the optimal nxn size for the best Gaussian blur filter. If the size is too large, would the filter be too blurred and produce an unusable image? What would be the best balance between using Gaussian blur to produce a smooth image while preserving the images detail. Also would a larger matrix size be computationally taxing?
jefforee
What is really interesting to observe is that the Gaussian Blur and Box Blur are not significantly different when looking from far away, but as we zoom we can see a more clear difference in their blurring capabilities.
I think Gaussian blur is a really cool technique that can be used to preprocess images. However I’m wondering what is the optimal nxn size for the best Gaussian blur filter. If the size is too large, would the filter be too blurred and produce an unusable image? What would be the best balance between using Gaussian blur to produce a smooth image while preserving the images detail. Also would a larger matrix size be computationally taxing?
What is really interesting to observe is that the Gaussian Blur and Box Blur are not significantly different when looking from far away, but as we zoom we can see a more clear difference in their blurring capabilities.