Hello
I'm having an issue calibrating a new light-field epifluorescence microscope we are constructing in the lab. It seems that the images behind the lenslets localized in the periphery of the microlens array go through some kind of blurring and stretching.
I'm not sure if it's due to optical aberration (coma maybe) or some how, they are out of focus.
Background image (illumination of FITC) is attached for illustration.
I'll be glad to get some insight about why it's happening and how to resolve this problem.
Thank you very much
Another image of trans-illumination (no specimen) to show the aberration at the peripheral lenslets
You can use a specific transformation that "transforms" back the image. This transformation is based on the model of distortions in the peripheral area of microscope which is due to specific blurring and geometric artifacts.
here is a citation that can help you in details:
VALDECASAS A. et al. On the extended depth of focus algorithms for bright field microscopy, MICRON, 2001, Vol. 32, Issue 6, pages 559-569
You can apply transform and/or models to reverse these artifacts or search for tje right algorithm.
Alternatively, you can try to us highly blurred (broad Gaussian filter) of the template and manipulate the image and then apply high pass filtering.
Thank you Viara for your answer.
I'm not sure these algorithms and kind of measurements are suitable for light-field images Moreover, changing the examine sample location (one of the procedures in the suggested algorithm) won't result in a better focus in the peripheral region - the problem is inherent in the microscope structure.
Right now I'm looking for hardware solutions to improve images quality.
Thank you
I am sure that you can search in the field to find how others have been solving the issue by using hardware. I guess, you should involve some additional optics but involving additional optics that is specific for the periphery may create other artifacts. So you may ask Google Patents and Google Scholar some search phrases and you be able to discover some reasonble way. Pls, let me know if you are getting closer. Wishing you luck with this one. But I am sure that a software approach by using some cylindrical transform may be easier to solve that problem.
- Badges
- Science topic
- Similar topics
- Computer Science
- Images