# How can estimate the PSF function to deblur an image of plate licence of a car [duplicate]

I want to deblur an image of a licence plate of a car. I have used the ImageDeconvolve[] function and Sharpen[] but the problem is still difficult because I need to estimate the kernel or the PSF function. A lot of techniques have been used to estimate the PSF. I wondered how the Fourier transform of the blurred image can give us an idea about the PSF function?

psfMotion[dim_?NumberQ, theta_?NumberQ] :=
Module[{PSF},
PSF = ConstantArray[0.0, {dim, dim}];
PSF[[Ceiling[dim / 2.0], All]] = 1.0 / dim;
Return[ImageRotate[Image[PSF], theta  Degree]];
];
psf = psfMotion[14, 0];
Table[ImageDeconvolve[licence, psfMotion[2, \[Theta]],
Method -> "RichardsonLucy", MaxIterations -> 15], {\[Theta], 0,
180}]

• In the movies, they just say: "computer: enhance!" If that doesn't work, it would be good to provide more detail in your question, e.g., an example image and what else you have tried.
– Jens
Commented Jul 9, 2014 at 19:09
• @Jens, I edited my question Commented Jul 9, 2014 at 20:36
• Did you try anything or are you simply hoping someone's either done this before or is really bored and will do it?
– acl
Commented Jul 9, 2014 at 20:58
• The example image is clearly limited by the image resolution (pixelation) and not by the point spread function of the camera. So it's pretty hopeless. There is a Wolfram Blog post about this topic.
– Jens
Commented Jul 9, 2014 at 22:58
• Let's enhance: youtube.com/watch?v=LhF_56SxrGk Commented Sep 18, 2015 at 6:57