Thresholding an image gives fuzzy black output

Question

Im preprocessing a qrcode to read it. I threshold the image, but sometimes I get a fuzzy black output, sometimes i get the proper binary image. Why is this happening? IM using C# and Emgucv.

Here is my code:

                    Image<Gray, byte> grayimg = inoutImg.Convert<Gray, byte>();  
                    grayimg = grayimg.ThresholdBinary(new Gray(this.tbbcrthreshold.Value), new Gray(255));
                    bcrthreshimgbox.Image = grayimg.Bitmap;

This is the Image im trying to threshold:

And this is the Image I get sometimes after i crop the image using roi :

Because of this, im unable to read the code.

How does your image look before thresholding? Are you displaying the image in a one to one scale, or is the image re-scaled? — JonasH, Mar 23 '22 at 07:39
Do we have to guess the original image type and content ? I am doing a computation and the answer is 553. What is going wrong ? — , Mar 23 '22 at 09:21
@YvesDaoust Bro, I have mentioned in the question that its a QR code image. 27.65/20 is 553. Atleast I tried to find your computational error! — Stormbringer, Mar 26 '22 at 06:26
I have updated the question with the QR code i was trying to threshold. — Stormbringer, Mar 26 '22 at 06:32
Oh, sorry, I did not mention that I was doing a computation the number 217. — , Mar 28 '22 at 08:04

score 0 · Answer 1 · answered Mar 28 '22 at 08:39

I tried with the qr code image you shared and works with every threshold value <200 (as it's the background value).

Here's an example with 100 as threashold.

 Image<Gray,byte> qr = new Image<Gray, byte>("qr.png");
 CvInvoke.Threshold(qr, qr, 100, 255, ThresholdType.Binary);
 CvInvoke.Imshow("qr",qr);

Output:

Thresholding an image gives fuzzy black output

1 Answers1