Filtration in frequency domain
Often the image is not an ideal and contains a lot of noise. This noise is represented as a high frequency in the frequency domain. Goal of this exercise will be to create a filter, which can remove the noise out of the image.
From the previous exercise we have completed DFT and IDFT which are key parts of the next program. Filtering in frequency domain can be done by simply multiplying certain frequency (most commonly in amplitude image) by a required value.
We can apply these types of filters:
- Low Pass
- High Pass
- Band Pass
We should also remind you, that from previous exercises the low frequencies are at the corners, so its necessary to switch first and third quadrant and second with forth.
- Quadrants switch
Example of low pass mask
Filtred Image
Sollution in LV
| Data | Value |
|---|---|
| Source | http://paulbourke.net/miscellaneous/imagefilter/https://homepages.inf.ed.ac.uk/rbf/HIPR2/fourier.htm |
| Code |