I have an image that is blurry:
This is part of the business card, and this is one of the frames made by the camera and without proper focus.
A clear image looks like this: 
I'm looking for a method that can give me a better quality image so that the image can be recognized by OCR, but should also be pretty fast. The image is not too blurry (I think so), but it is not good for OCR. I tried:
- various types of HPF,
- Laplacian
- Landing detector,
- combinations of morphological operations (opening, closing).
I also tried:
- deconvolution with a Wiener filter,
- deconvolution and the Lucy-Richardson method.
But finding a PSF (point spread function) was not easy. This method is considered effective, but not so fast. I also tried FFT and then IFFT with a Gaussian mask, but the results were unsatisfactory. I am looking for some general method of deforming images with text, and not just this image. Can someone help me with this problem? I would be grateful for any advice. I work with OpenCV 3 (C ++ and sometimes Python).
Artur source share