Conclusions
The ultimate goal of the restoration techniques is to
improve the visual-quality of a degraded image. In this
paper, we focus on the discussion of the degradation
models and the restoration approaches–the Inverse filter
and the Wiener Filter. We compare the performance with
the synthetic lena and vehicle number plate images.
Then, we show a real case solved by the restoration technique.
Because the noise effect in the real case is not serious,
the restoration results by using the inverse filter and
the Wiener filter are similar. In practical applications,
since the parameters of the degraded processes are not
easy to be obtained exactly, we suggest the first step is
trying the Inverse filter, since it is simple. In future work,
we will try to compare the observed (blurred) images
with the ideal distorted digits (by known PSF’s), and see
if we can obtain better recognition.