We approach the removal of back-to-front interferences from recto and verso scans of archival documents as a blind source separation problem, considering the front and back ideal images as two individual patterns that overlap in the observed scans through some mixing operator. The nonlinear mixing model and the related restoration algorithm proposed in [1] are efficient for modern documents affected by mild show-through, but are not fully adequate to cope with ancient documents often degraded by the heavier and non-stationary bleed-through distortion. We then propose to modify this data model to account for non-stationarity of the degradation, and resort to the genuine concept of source separation for deriving the restoration algorithm. Within a regularization approach, we joint estimate the ideal images and the model parameters, by minimizing an energy function of all the unknowns, accounting also for local autocorrelation of the the ideal images. We derive a fully deterministic algorithm that is computationally efficient, and analyze its performance against documents heavily degraded by either show-through or bleed-through.

Restoration of recto-verso archival documents through a regularized nonlinear model

GERACE, Ivan;MARTINELLI, FRANCESCA;
2012

Abstract

We approach the removal of back-to-front interferences from recto and verso scans of archival documents as a blind source separation problem, considering the front and back ideal images as two individual patterns that overlap in the observed scans through some mixing operator. The nonlinear mixing model and the related restoration algorithm proposed in [1] are efficient for modern documents affected by mild show-through, but are not fully adequate to cope with ancient documents often degraded by the heavier and non-stationary bleed-through distortion. We then propose to modify this data model to account for non-stationarity of the degradation, and resort to the genuine concept of source separation for deriving the restoration algorithm. Within a regularization approach, we joint estimate the ideal images and the model parameters, by minimizing an energy function of all the unknowns, accounting also for local autocorrelation of the the ideal images. We derive a fully deterministic algorithm that is computationally efficient, and analyze its performance against documents heavily degraded by either show-through or bleed-through.
2012
978-1-4673-1068-0
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11391/1225293
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 4
  • ???jsp.display-item.citation.isi??? 3
social impact