Please use this identifier to cite or link to this item: http://hdl.handle.net/10316/43904
Title: Analysis of adaptive forward-backward diffusion flows with applications in image processing
Authors: Prasath, V. B. Surya 
Urbano, José Miguel 
Vorotnikov, Dmitry 
Issue Date: 2015
Publisher: IOP Publishing
Project: PEst-C/MAT/UI0324/2011 
Serial title, monograph or event: Inverse Problems
Volume: 31
Issue: 10
Abstract: The nonlinear diffusion model introduced by Perona and Malik (1990 IEEE Trans. Pattern Anal. Mach. Intell. 12 629–39) is well suited to preserve salient edges while restoring noisy images. This model overcomes well-known edge smearing effects of the heat equation by using a gradient dependent diffusion function. Despite providing better denoizing results, the analysis of the PM scheme is difficult due to the forward-backward nature of the diffusion flow. We study a related adaptive forward-backward diffusion equation which uses a mollified inverse gradient term engrafted in the diffusion term of a general nonlinear parabolic equation. We prove a series of existence, uniqueness and regularity results for viscosity, weak and dissipative solutions for such forward-backward diffusion flows. In particular, we introduce a novel functional framework for wellposedness of flows of total variation type. A set of synthetic and real image processing examples are used to illustrate the properties and advantages of the proposed adaptive forward-backward diffusion flows.
URI: http://hdl.handle.net/10316/43904
DOI: 10.1088/0266-5611/31/10/105008
10.1088/0266-5611/31/10/105008
Rights: embargoedAccess
Appears in Collections:I&D CMUC - Artigos em Revistas Internacionais

Files in This Item:
File Description SizeFormat
wpde_revised2.pdf8.09 MBAdobe PDFView/Open
Show full item record

SCOPUSTM   
Citations

14
checked on Nov 9, 2022

WEB OF SCIENCETM
Citations 5

13
checked on Jul 2, 2022

Page view(s) 50

369
checked on Nov 28, 2022

Download(s)

281
checked on Nov 28, 2022

Google ScholarTM

Check

Altmetric

Altmetric


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.