Diffusion-based enhancement methods

Diffusion-based and PDE-based methods are powerful tools in image processing. These methods are extremely local, and therefore are very suitable for problems where there is a high degree spatial correlation. for example in the field of document image processing where many of physical phenomena which occur to the paper and ink can be easily described using some simple diffusion-based models. However, direct application of diffusion-based methods to the degraded historical document images is not very successful. The main reason is that several layers of information, such as degraded background and interfering signatures, are added to the document view through "non-local" processes such as ink seepage and external alteration.

In our novel diffusion-based approach to enhancment of degraded document images, the model is generalized in such a way that diffusion processes between various sources of information are allowed. Not only this generalization enables us to provide promising degradation modeling (link), but also it provides the required non-locality in the enhancement method.

For more information, please refer to [1].

A schematic diagram of a generalized diffusion-based model for the case of doubled-sided documents. The reverse diffusion of the verso side image helps in the removal of interference patterns:

Schema diffusionSchema diffusion


References

ericssonlogo
inocybelogo
canalogo
cienalogo
Civimetrix Telecom logo
mitacslogo
risq logo
nserclogo
promptlogo
ecolepolytechniquelogo
University of Torontologo
frqntlogo
uqlogo
MDEIE logo
cfilogo
ciraiglogo