This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Image Processing techniques to remove diffraction Artifacts

Hi

I am trying to create a programmable aperture using a DLP but due to the nature of the application I appear to get lots of diffraction artifacts in my images (see attached).

Are there any image processing techniques that I can use to remove these artifacts. i.e. post processing techniques.

In the images below we see a picture of a dot. Only 1 dot is being photographed (bottom right) but 3 other dots appear around it. Can this be removed post processing?

It was suggested to Use 2D FIR filters. Not sure if any other techniques are possible?

  • Hello Ashill,

    Are you using any imaging lens to form the image? 

    Using an imaging lens just as you would with a broad band illumination to form the image will also re-converge the image of the dot into one.

    Fizix

  • Hi

    Yes i am using an imaging lens to form the image.
  • Hello again Ashill,

    Thank you.  Would it be possible to make a simple diagram of your optical system from illumination source to image?  It would be helpful.

    Fizix

  • Hi fizix

    Please see diagram above. Sometimes if the object is far enough away the light is already collimated so i do omit lens 1. The objective is basically to get the dlp into the aperture plane.

    Thank you

  • Hi,

    Things will become more simple if you place the DMD plane parallel to the image plane.

    If that is not feasible for certain reasons, you shall take advantage of the   https://en.wikipedia.org/wiki/Scheimpflug_principle

  • Ashill,

    Your arrangement shows the DMD in collimated space which puts the DMD surface at a Fourier transform plane of the object.  This WILL give you diffraction artifacts.  I know of no way to remove them in this scheme.

    Here is what I was implying (see diagram):

    Of course this may defeat what you are attempting to do.  Otherwise you will have to move far enough away  from the DMD that you isolate only one order in your current configuration.  Specifically only one order gets into the lens or you put an aperture on the outbound lens to only capture one order.

    I hope this helps.

    P.S.  To DLP_ViALUX - I think his diagram was not intended to represent relative parallel surfaces.  In any case Sheimpflug will not solve the diffraction artifacts if the DMD is in collimated space.

  • Hi Fizix

    Thank you very much for the reply and diagram. I really appreciate it.

    With regards to the configuration you suggested, yes it is defeating the purpose of my application as I do not wish to form an image on the DLP. I have been told that the only way to remove the diffraction is post processing techniques such as FIR filters. Is there any other way using image processing techniques that I can use to remove the artifacts. 

  • Hello Ashill,

    I am not aware of any post processing techniques to do it, but that does not mean they do not exist.

    Is using an iris at the second lens to restrict the rays to those only from the brightest order not an option?

    Fizix

  • Hi Fizix

    It most certainly is an option. I just wanted to check if other post processing techniques exist before i introduce another aperture into the system. Hopefully that is able to help with the diffraction effects.

    Regards

    Ashill

  • Ashill,

    I am not aware of any post processing techniques, but you may want to search some of the literature from the MOEMS conferences at SPIE. And other research done on DMDs.