We propose a two-stage variational model for the additive decomposition of images into piecewise constant, smooth, textured and white noise components. The challenging separation of noise from texture is successfully achieved by including a normalized whiteness constraint in the model, and the selection of the regularization parameters is performed based on a novel multi-parameter cross-correlation principle. The two resulting minimization problems are efficiently solved by means of the alternating directions method of multipliers. Numerical results show the potentiality of the proposed model for the decomposition of textured images corrupted by several kinds of additive white noises.
Quaternary Image Decomposition with Cross-Correlation-Based Multi-parameter Selection / Girometti L.; Huska M.; Lanza A.; Morigi S.. - STAMPA. - 14009:(2023), pp. 120-133. (Intervento presentato al convegno 9th International Conference on Scale Space and Variational Methods in Computer Vision, SSVM 2023 tenutosi a ita nel 2023) [10.1007/978-3-031-31975-4_10].
Quaternary Image Decomposition with Cross-Correlation-Based Multi-parameter Selection
Girometti L.;Huska M.;Lanza A.;Morigi S.
2023
Abstract
We propose a two-stage variational model for the additive decomposition of images into piecewise constant, smooth, textured and white noise components. The challenging separation of noise from texture is successfully achieved by including a normalized whiteness constraint in the model, and the selection of the regularization parameters is performed based on a novel multi-parameter cross-correlation principle. The two resulting minimization problems are efficiently solved by means of the alternating directions method of multipliers. Numerical results show the potentiality of the proposed model for the decomposition of textured images corrupted by several kinds of additive white noises.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
