Ask Your Question
0

OpenCV - Gaussian Noise

asked 2014-07-04 18:24:18 -0600

JoeMama gravatar image

Hello,

here's my problem: I'm trying to create a simple program which adds Gaussian noise to an input image. The only constraints are that the input image is of type CV_64F (i.e. double) and the values are and must be kept normalized between 0 and 1.

The code I wrote is the following:

Mat my_noise;
my_ noise = Mat (input.size(), input.type());

randn(noise, 0, 5); //mean and variance

input += noise;

The above code doesn't work, the resulting image doesn't get displayed properly. I think that happens because it gets out of the 0,1 range. I modified the code like this:

Mat my_noise;
my_ noise = Mat (input.size(), input.type());

randn(noise, 0, 5); //mean and variance

input += noise;

normalize(input, input, 0.0, 1.0, CV_MINMAX, CV_64F);

but it still doesn't work. Again, the resulting image doesn't get displayed properly. Where is the problem? Remember: the input image is of type CV_64F and the values are normalized between 0 and 1 before adding noise and have to remain like also after the noise addition.

Thank you in advance.

edit retag flag offensive close merge delete

Comments

1

Why not using cv::GaussinBlur() instead?

thdrksdfthmn gravatar imagethdrksdfthmn ( 2014-07-07 03:52:09 -0600 )edit
2

@thdrksdfthmn because he wants to add noise, not apply blurring to remove possible noise. You should read with attention ;)

StevenPuttemans gravatar imageStevenPuttemans ( 2014-07-07 06:43:06 -0600 )edit

2 answers

Sort by ยป oldest newest most voted
5

answered 2014-07-07 02:48:18 -0600

updated 2014-07-07 02:49:51 -0600

There may be two reasons for the results you got:

  • The noise matrix and the image are not normalized in the same range of [0,1].
  • The stddev value you used (5) was too large.

The code following works for me:

// imGray is the grayscale of the input image
cv::Mat noise = Mat(imGray.size(),CV_64F);
normalize(imGray, result, 0.0, 1.0, CV_MINMAX, CV_64F);
cv::randn(noise, 0, 0.05);
result = result + noise;
normalize(result, result, 0.0, 1.0, CV_MINMAX, CV_64F);
cv::imshow("OUTPUT",result);

And here is the illustration (an input image and Gaussian noise version with stddev=0.05 and 0.1, respectively):

image descriptionimage descriptionimage description

edit flag offensive delete link more
0

answered 2015-02-04 06:57:22 -0600

    Mat input = imread("inputImage.png");
    Mat imGray;
    cvtColor(input,imGray,CV_BGR2GRAY);
    cv::Mat noise = Mat(imGray.size(),CV_64F);
    Mat result;
    normalize(imGray, result, 0.0, 1.0, CV_MINMAX, CV_64F);
    cv::randn(noise, 0, 0.1);
    result = result + noise;
    normalize(result, result, 0.0, 1.0, CV_MINMAX, CV_64F);
    result.convertTo(result, CV_32F, 255, 0);
    cv::imwrite("result.png",result);
    cv::waitKey(0);
edit flag offensive delete link more

Question Tools

Stats

Asked: 2014-07-04 18:24:18 -0600

Seen: 29,042 times

Last updated: Jul 07 '14