|
Last updated on 5
Jan. 2005 |
The commercial project made for presentation of new algorithm of deblurring
of raster images. All rights on application of the given algorithm belong
to the author of the project. |
||
|
Deblurring Denoise Deconvolution Dequantization |
DB-M-Spline - the new, linear, best quality and fast algorithm of image deblurring.
Deblurring - to recover a
sharp image from its blurry observation. It frequently arises in imaging
sciences and technologies, including optical, medical, and astronomical
applications, and is a crucial step towards successful detections of
important patterns such as abnormal tissues and surface details of a distant
planet. Mathematically, image deblurring is intimately connected to back-ward
diffusion processes (e.g., inverting the heat equation), which are
notoriously unstable. As inverse problem solvers, deblurring models therefore
crucially depend upon proper regularizers or conditioners that help secure
stability, often at the necessary cost of losing certain high-frequency
details in original clear images. Such regularization techniques can ensure
the existence or uniqueness of deblurred images. The main
advantage of these methods is high speed of the realization. However, their
application leads to a strong artifacts which, by permanently increasing
demands for quality of interpolation, significantly limits the area of their
application. Recently, on the market of
professional graphic programmes, quite a few more effective and better
quality algorithms have appeared. Along with the linear methods (inverse filter) such as Wiener filter, Unsharp Mask (Photoshop, Focus Magic)а non linear
iterations (or adaptive) algorithms appeared Ц Lucy-Richardson
algorithm, blind deconvolution
algorithm etc. The
given methods allow to receive good results, however they are still far from
perfect. The aim of data spelling appears as the
presentationа of the new linear method
which allows to carry out scaling of a better quality, better than all known
methods, both linear and adaptive. The main advantage of suggested method
over the rest of them is the speed (because it appears linear) simpleness of
its realization and best quality closest to achievable limits. The method
brought to your attention was called DeBlurring Magic spline or DB-M-spline
by its author. DB-M-spline - algorithm is fully worked out and realized
exclusively by the author and all the rights of its use until the decision of
sale, belong to the author. The base of the suggested algorithm (DB-M-spline) lies in application of the new revolutionary approach at working out the algorithm. Valuating the form and size of approximating function, it is also possible to ascertain that obtained algorithms allow getting the quality of the deblurring which is maximally approximate to the achievable limit from the theory of information point of view. There are also a few algorithm modifications - DB-M-spline-33, DB-M-spline-35, DB-M-spline-53, DB-M-spline-55. In order to confirm above, I suggest to get
familiar with practical realization of the algorithm on the well known
examples of tasks and solutions. As the
test the following representation has been chosen: ааааа
lhouseаа аааааааааааааааааааааааааааааааааааааflowersаааааааааааааааааааааааааааааааааа lhouse2 As the
real image test the following representation has been chosen:
To see the picture click on the fragment Test - flowers Original image - it will be blurredааааааа
Blurred image + noise
Deblurred image: аааааа
ааааWiener filter ааааааааааааааLucy-Richardsonаааааааааааа Unsharp maskаааааааааааааа DB-MSpline-55 аааа а(New method)
Test - lhouse Original low-resolution image - it will be enlarged by 3 times
Blurred image + noise
аааааа
ааааWiener filter ааааааааааааааLucy-Richardsonаааааааааааа Unsharp maskаааааааааааааа DB-MSpline-55 аааааа (New method)
Test - lhouse2 Original low-resolution image - it will be enlarged by 3 times
Blurred image + noise
ааааааааа Wiener filter ааааааааааааааLucy-Richardsonаааааааааааа Unsharp maskаааааааааааааааа DB-MSpline-55 (New method)
TEST of REAL IMAGE Even new professional digital chambers are not free from problems with
discrepancy of the optical resolution of received pictures physical
resolution (to quantity of pixels CCD matrix). Even at one of the new camera
Minolta D7 is have of the image degradation. These are distortions arise by
virtue of lines of the reasons and cannot be removed completely. Noticeably
to improve value judgment of quality of perception of existing pictures and
to raise the optical resolution the offered method will allow. You can
estimate work of algorithm in real photos made by real digital devices. Unsharp
maskааааааааааааа Real image before
deblurringааааааааааааа DB-MSpline-55
Unsharp
maskааааааааааааа Real image before
deblurringааааааааааааа DB-MSpline-55
Unsharp
maskаааааааааааааа Real image before
deblurringааааааааааааа DB-MSpline-55
Unsharp
maskаааааааааааааа Real image before
deblurringааааааааааааа DB-MSpline-55
Minolta 7D Unsharp
maskаааааааааааааа Real image before
deblurringааааааааааааа DB-MSpline-55
Minolta 7D Unsharp
mask ааааааааааааааReal image before
deblurringааааааааааааа DB-MSpline-55
Minolta 7D Unsharp
maskаааааааааааааа Real image before
deblurringааааааааааааа DB-MSpline-55
Minolta 7D Unsharp
maskаааааааааааааа Real image before
deblurringааааааааааааа DB-MSpline-55
vitalybn@fpy.ru resampling@yandex.ru |
