2
$\begingroup$

I have c# desktop application and I am using emgu framework as a wrapper for OpenCV.

I am trying to sharpen an image because the input images are not great.

I tired a technique I found via searches that suggest from the original image:

I take a copy and blur it. I then take another copy of the original and invert it.
I then merge the 2 new images together.
I then finally invert that merged image.

I have done all that but the image pretty much looks the same.

Is this the right approach?

I can post my code if need be - I used Aforge to quickly test this approach though.

UPDATE: To answer my own question I used this:

            Image<Bgr, byte> image = new Image<Bgr, byte>(@"D:\20140320022038047.jpg");
            Image<Bgr, byte> image2 = new Image<Bgr, byte>(@"D:\20140320022038047.jpg");
            Emgu.CV.CvInvoke.cvSmooth(image, image2, SMOOTH_TYPE.CV_GAUSSIAN, 5, 5, 9, 9);
            Emgu.CV.CvInvoke.cvAddWeighted(image, 1.5, image2, -0.5, 0, image);
$\endgroup$
3
$\begingroup$

I know this answer is very late but I think this question worth an answer

The algorithm you used goes as follows:

  1. Blur the original image using the Gaussian filter with given Mask Size & Sigma

  2. Subtract the blurred image from the original (result is called Mask) to eliminate background and get the edges regions

  3. Add a weighted portion from the mask to the original image by multiplying the Mask (the edges only) by K to enhance edges regions

K: User input (If K = 1 Unsharp, If K > 1 Highboost)

Mask(x,y) = Orig(x,y) – Blurred(x,y)

Result(x,y) = Orig(x,y) + K × Mask(x,y)

Here is the code:

public static Image<Gray, byte> Sharpen(Image<Gray, byte> image, int w, int h, double sigma1, double sigma2, int k)
    {
        w = (w % 2 == 0) ? w - 1 : w;
        h = (h % 2 == 0) ? h - 1 : h;
        //apply gaussian smoothing using w, h and sigma 
        var gaussianSmooth = image.SmoothGaussian(w, h, sigma1, sigma2);
        //obtain the mask by subtracting the gaussian smoothed image from the original one 
        var mask = image - gaussianSmooth;
        //add a weighted value k to the obtained mask 
        mask *= k;
        //sum with the original image 
        image += mask;
        return image;
    }

I think your problem is that you didn't subtract the smoothed image from the original image, anyway I hope this help

| improve this answer | |
$\endgroup$
  • $\begingroup$ worth giving a go. Which I will as soon as I get time this evening. Thanks :) $\endgroup$ – Andrew Simpson Feb 3 '15 at 16:27

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.