- Carefully observe it, if the involved noise is familiar (like impulse noise, salt and pepper, some casse gaussian too), you may predict it directly by observing the images.
- If not (Step 1), See the Historam of the image, from the distrbution map of the histogram, you may get clue about the type of noise involved on images, afterwards do estimation of noise type. For better understanding, please refer the Gonzalez Book (Chapters Image Enghancement or Image Restoration) or this PPT directly
- If you get the idea, create an approprite h using fspecial
- Then proceed for imfilter with h, defined in step 3.
- Some cases, you may have check with multiple h (custom also), see the results, where you get the best results.
Face image preprocesaing filtering
2 Ansichten (letzte 30 Tage)
Ältere Kommentare anzeigen
I want to filter face image.
How to filter image with mathematical calculations?
How to remove blur, noise in face area?
How to denoise image?
0 Kommentare
Antworten (2)
KALYAN ACHARJYA
am 12 Apr. 2020
You did not attach the sample image. The question is quite broad, it is quit unfair to answer image processing noise related question without observing the noisy image.
Here are some suggestions-
Good Wishes!
Image Analyst
am 12 Apr. 2020
Bearbeitet: Image Analyst
am 12 Apr. 2020
Like Kalyan said, really broad question. We can't give a whole course here, so I'll just point you to some functions:
Deblur: conv2(), imfilter(), deconvlucy()
Denoise: medfilt2(), imfilter(), imnlmfilt() (non local means), conv2(), imbilatfilt(), imdiffusefilt(), etc.
Also see my face pixelating/blurring script in this link
3 Kommentare
Image Analyst
am 13 Apr. 2020
Taking the histogram just describes the image - it does not enhance it, though the histogram might be used in deciding some parameters of the enhancement algorithm.
Again, enhancement is also such a broad topic. You could do a simple contrast stretch with imadjust(), or you could do something like adapthisteq(). If you want more algorithms, you could browse the papers here: Vision Bibliography
Siehe auch
Produkte
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!