Deep Inertia $L_p$ Half-Quadratic Splitting Unrolling Network for Sparse View CT Reconstruction

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Guo, Yu, Wu, Caiying, Li, Yaxin, Jin, Qiyu, Zeng, Tieyong
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866913466820853760
author Guo, Yu
Wu, Caiying
Li, Yaxin
Jin, Qiyu
Zeng, Tieyong
author_facet Guo, Yu
Wu, Caiying
Li, Yaxin
Jin, Qiyu
Zeng, Tieyong
contents Sparse view computed tomography (CT) reconstruction poses a challenging ill-posed inverse problem, necessitating effective regularization techniques. In this letter, we employ $L_p$-norm ($0<p<1$) regularization to induce sparsity and introduce inertial steps, leading to the development of the inertial $L_p$-norm half-quadratic splitting algorithm. We rigorously prove the convergence of this algorithm. Furthermore, we leverage deep learning to initialize the conjugate gradient method, resulting in a deep unrolling network with theoretical guarantees. Our extensive numerical experiments demonstrate that our proposed algorithm surpasses existing methods, particularly excelling in fewer scanned views and complex noise conditions.
format Preprint
id arxiv_https___arxiv_org_abs_2408_06600
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Deep Inertia $L_p$ Half-Quadratic Splitting Unrolling Network for Sparse View CT Reconstruction
Guo, Yu
Wu, Caiying
Li, Yaxin
Jin, Qiyu
Zeng, Tieyong
Image and Video Processing
Computer Vision and Pattern Recognition
Sparse view computed tomography (CT) reconstruction poses a challenging ill-posed inverse problem, necessitating effective regularization techniques. In this letter, we employ $L_p$-norm ($0<p<1$) regularization to induce sparsity and introduce inertial steps, leading to the development of the inertial $L_p$-norm half-quadratic splitting algorithm. We rigorously prove the convergence of this algorithm. Furthermore, we leverage deep learning to initialize the conjugate gradient method, resulting in a deep unrolling network with theoretical guarantees. Our extensive numerical experiments demonstrate that our proposed algorithm surpasses existing methods, particularly excelling in fewer scanned views and complex noise conditions.
title Deep Inertia $L_p$ Half-Quadratic Splitting Unrolling Network for Sparse View CT Reconstruction
topic Image and Video Processing
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2408.06600