Abstract : In this work we introduce a novel operator $$\displaystyle \varDelta _{(p,q)}$$ as an extended family of operators that generalize the p-Laplace operator. The operator is derived with an emphasis on image processing applications, and particularly, with a focus on image denoising applications. We propose a non-linear transition function, coupling p and q, which yields a non-linear filtering scheme analogous to adaptive spatially dependent total variation and linear filtering. Well-posedness of the final parabolic PDE is established via pertubation theory and connection to classical results in functional analysis. Numerical results demonstrates the applicability of the novel operator $$\displaystyle \varDelta _{(p,q)}$$.
Lorena Bociu; Jean-Antoine Désidéri; Abderrahmane Habbal. 27th IFIP Conference on System Modeling and Optimization (CSMO), Jun 2015, Sophia Antipolis, France. Springer International Publishing, IFIP Advances in Information and Communication Technology, AICT-494, pp.107-116, 2016, System Modeling and Optimization. 〈10.1007/978-3-319-55795-3_9〉
https://hal.inria.fr/hal-01626927
Contributeur : Hal Ifip
<>
Soumis le : mardi 31 octobre 2017 - 14:41:55
Dernière modification le : mardi 31 octobre 2017 - 14:44:49
Document(s) archivé(s) le : jeudi 1 février 2018 - 13:42:40
George Baravdish, Yuanji Cheng, Olof Svensson, Freddie Åström. Extension of p-Laplace Operator for Image Denoising. Lorena Bociu; Jean-Antoine Désidéri; Abderrahmane Habbal. 27th IFIP Conference on System Modeling and Optimization (CSMO), Jun 2015, Sophia Antipolis, France. Springer International Publishing, IFIP Advances in Information and Communication Technology, AICT-494, pp.107-116, 2016, System Modeling and Optimization. 〈10.1007/978-3-319-55795-3_9〉. 〈hal-01626927〉