Ask Your Question
0

How to identify pixels with noise ?

asked 2016-12-01 20:29:17 -0600

rugby2312 gravatar image

updated 2016-12-01 20:31:11 -0600

Hi Everyone,

I was wondering if there is any function for detecting (just detecting, not correcting) the number of pixels with noise within an image

Take this picture for example :http://imgur.com/VR7bTWV , whose size is 215x215 (or 46,225 pixels).

Would it be possible to get the total number of noisy pixels ? i.e 30,000 noisy pixels out of 46,225 total pixels.

My idea is to apply noise filtering only if there is a significant amount of noise as noise filtering usually reduces details in the image, also I want to apply different filtering levels based on the existing noise levels .

What approach do you suggest ?

Thanks in Advance.

edit retag flag offensive close merge delete

1 answer

Sort by ยป oldest newest most voted
1

answered 2016-12-01 20:53:39 -0600

Tetragramm gravatar image

Unfortunately, all of the pixels in that image have noise. Typical image noise is what is called Gaussian noise. The noise follows a Gaussian distribution. So all pixels have noise, but most have amounts close to zero. The wider the spread of values (the wider the gaussian distribution) the more noisy an image appears. Most work focuses on estimating the width of the gaussian distribution, properly called the standard deviation.

HERE is a paper on estimating noise, and there are surely more a google search away.

edit flag offensive delete link more

Comments

Thanks , I'll take a look at it.

rugby2312 gravatar imagerugby2312 ( 2016-12-01 21:42:20 -0600 )edit

Question Tools

1 follower

Stats

Asked: 2016-12-01 20:29:17 -0600

Seen: 126 times

Last updated: Dec 01 '16