IEEE Transactions on Computational Imaging最新文献_第3页

FoveaSPAD: Exploiting Depth Priors for Adaptive and Efficient Single-Photon 3D Imaging FoveaSPAD：利用深度先验进行自适应和高效的单光子3D成像

IF 4.2 2区计算机科学

IEEE Transactions on Computational Imaging Pub Date : 2024-12-09 DOI: 10.1109/TCI.2024.3503360

Justin Folden;Atul Ingle;Sanjeev J. Koppal

{"title":"FoveaSPAD: Exploiting Depth Priors for Adaptive and Efficient Single-Photon 3D Imaging","authors":"Justin Folden;Atul Ingle;Sanjeev J. Koppal","doi":"10.1109/TCI.2024.3503360","DOIUrl":"https://doi.org/10.1109/TCI.2024.3503360","url":null,"abstract":"Fast, efficient, and accurate depth-sensing is important for safety-critical applications such as autonomous vehicles. Direct time-of-flight LiDAR has the potential to fulfill these demands, thanks to its ability to provide high-precision depth measurements at long standoff distances. While conventional LiDAR relies on avalanche photodiodes (APDs), single-photon avalanche diodes (SPADs) are an emerging image-sensing technology that offer many advantages such as extreme sensitivity and time resolution. In this paper, we remove the key challenges to widespread adoption of SPAD-based LiDARs: their susceptibility to ambient light and the large amount of raw photon data that must be processed to obtain in-pixel depth estimates. We propose new algorithms and sensing policies that improve signal-to-noise ratio (SNR) and increase computing and memory efficiency for SPAD-based LiDARs. During capture, we use external signals to \u0000<italic>foveate</i>\u0000, i.e., guide how the SPAD system estimates scene depths. This foveated approach allows our method to “zoom into” the signal of interest, reducing the amount of raw photon data that needs to be stored and transferred from the SPAD sensor, while also improving resilience to ambient light. We show results both in simulation and also with real hardware emulation, with specific implementations achieving a 1548-fold reduction in memory usage, and our algorithms can be applied to newly available and future SPAD arrays.","PeriodicalId":56022,"journal":{"name":"IEEE Transactions on Computational Imaging","volume":"10 ","pages":"1722-1735"},"PeriodicalIF":4.2,"publicationDate":"2024-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142810335","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Light Field Angular Super-Resolution Network Based on Convolutional Transformer and Deep Deblurring 基于卷积变压器和深度去模糊的光场角超分辨网络

IF 4.2 2区计算机科学

IEEE Transactions on Computational Imaging Pub Date : 2024-12-09 DOI: 10.1109/TCI.2024.3507634

Deyang Liu;Yifan Mao;Yifan Zuo;Ping An;Yuming Fang

{"title":"Light Field Angular Super-Resolution Network Based on Convolutional Transformer and Deep Deblurring","authors":"Deyang Liu;Yifan Mao;Yifan Zuo;Ping An;Yuming Fang","doi":"10.1109/TCI.2024.3507634","DOIUrl":"https://doi.org/10.1109/TCI.2024.3507634","url":null,"abstract":"Many Light Field (LF) angular super-resolution methods have been proposed to cope with the LF spatial and angular resolution trade-off problem. However, most existing methods cannot simultaneously explore LF local and non-local geometric information, which limits their performances. Moreover, since the quality degradation model of the reconstructed dense LF is always neglected, most solutions fail to effectively suppress the blurry edges and artifacts. To overcome these limitations, this paper proposes an LF angular super-resolution network based on convolutional Transformer and deep deblurring. The proposed method mainly comprises a Global-Local coupled Convolutional Transformer Network (GLCTNet), a Deep Deblurring Network (DDNet), and a Texture-aware feature Fusion Network (TFNet). The GLCTNet can fully capture the long-range dependencies while strengthening the locality of each view. The DDNet is utilized to construct the quality degradation model of the reconstructed dense LF to suppress the introduced blurred edges and artifacts. The TFNet distills the texture features by extracting the local binary pattern map and gradient map, and allows a sufficient interaction of the obtained non-local geometric information, local structural information, and texture information for LF angular super-resolution. Comprehensive experiments demonstrate the superiority of our proposed method in various LF angular super-resolution tasks. The depth estimation application further verifies the effectiveness of our method in generating high-quality dense LF.","PeriodicalId":56022,"journal":{"name":"IEEE Transactions on Computational Imaging","volume":"10 ","pages":"1736-1748"},"PeriodicalIF":4.2,"publicationDate":"2024-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142810336","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Low-Rank Angular Prior Guided Multi-Diffusion Model for Few-Shot Low-Dose CT Reconstruction 用于少镜头低剂量 CT 重建的低秩角度先验引导多扩散模型

IF 4.2 2区计算机科学

IEEE Transactions on Computational Imaging Pub Date : 2024-12-04 DOI: 10.1109/TCI.2024.3503366

Wenhao Zhang;Bin Huang;Shuyue Chen;Xiaoling Xu;Weiwen Wu;Qiegen Liu

{"title":"Low-Rank Angular Prior Guided Multi-Diffusion Model for Few-Shot Low-Dose CT Reconstruction","authors":"Wenhao Zhang;Bin Huang;Shuyue Chen;Xiaoling Xu;Weiwen Wu;Qiegen Liu","doi":"10.1109/TCI.2024.3503366","DOIUrl":"https://doi.org/10.1109/TCI.2024.3503366","url":null,"abstract":"Low-dose computed tomography (LDCT) is essential in clinical settings to minimize radiation exposure; however, reducing the dose often leads to a significant decline in image quality. Additionally, conventional deep learning approaches typically require large datasets, raising concerns about privacy, costs, and time constraints. To address these challenges, a few-shot low-dose CT reconstruction method is proposed, utilizing low-Rank Angular Prior (RAP) multi-diffusion model. In the prior learning phase, projection data is transformed into multiple consecutive views organized by angular segmentation, allowing for the extraction of rich prior information through low-rank processing. This structured approach enhances the learning capacity of the multi-diffusion model. During the iterative reconstruction phase, a stochastic differential equation solver is employed alongside data consistency constraints to iteratively refine the acquired projection data. Furthermore, penalized weighted least-squares and total variation techniques are integrated to improve image quality. Results demonstrate that the reconstructed images closely resemble those obtained from normal-dose CT, validating the RAP model as an effective and practical solution for artifact and noise reduction while preserving image fidelity in low-dose situation.","PeriodicalId":56022,"journal":{"name":"IEEE Transactions on Computational Imaging","volume":"10 ","pages":"1763-1774"},"PeriodicalIF":4.2,"publicationDate":"2024-12-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142821247","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

MVMS-RCN: A Dual-Domain Unified CT Reconstruction With Multi-Sparse-View and Multi-Scale Refinement-Correction MVMS-RCN：一种多稀疏视图和多尺度精细校正的双域统一CT重建方法

IF 4.2 2区计算机科学

IEEE Transactions on Computational Imaging Pub Date : 2024-11-27 DOI: 10.1109/TCI.2024.3507645

Xiaohong Fan;Ke Chen;Huaming Yi;Yin Yang;Jianping Zhang

{"title":"MVMS-RCN: A Dual-Domain Unified CT Reconstruction With Multi-Sparse-View and Multi-Scale Refinement-Correction","authors":"Xiaohong Fan;Ke Chen;Huaming Yi;Yin Yang;Jianping Zhang","doi":"10.1109/TCI.2024.3507645","DOIUrl":"https://doi.org/10.1109/TCI.2024.3507645","url":null,"abstract":"X-ray Computed Tomography (CT) is one of the most important diagnostic imaging techniques in clinical applications. Sparse-view CT imaging reduces the number of projection views to a lower radiation dose and alleviates the potential risk of radiation exposure. Most existing deep learning (DL) and deep unfolding sparse-view CT reconstruction methods: 1) do not fully use the projection data; 2) do not always link their architecture designs to a mathematical theory; 3) do not flexibly deal with multi-sparse-view reconstruction assignments. This paper aims to use mathematical ideas and design optimal DL imaging algorithms for sparse-view CT reconstructions. We propose a novel dual-domain unified framework that offers a great deal of flexibility for multi-sparse-view CT reconstruction through a single model. This framework combines the theoretical advantages of model-based methods with the superior reconstruction performance of DL-based methods, resulting in the expected generalizability of DL. We propose a refinement module that utilizes unfolding projection domain to refine full-sparse-view projection errors, as well as an image domain correction module that distills multi-scale geometric error corrections to reconstruct sparse-view CT. This provides us with a new way to explore the potential of projection information and a new perspective on designing network architectures. The multi-scale geometric correction module is end-to-end learnable, and our method could function as a plug-and-play reconstruction technique, adaptable to various applications. Extensive experiments demonstrate that our framework is superior to other existing state-of-the-art methods.","PeriodicalId":56022,"journal":{"name":"IEEE Transactions on Computational Imaging","volume":"10 ","pages":"1749-1762"},"PeriodicalIF":4.2,"publicationDate":"2024-11-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142810334","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

An Adaptive Photon Count Reconstruction Algorithm for Sparse Count and Strong Noise Count Data With Low Signal Background Ratio 低信号背景比下稀疏计数和强噪声计数数据的自适应光子计数重建算法

IF 4.2 2区计算机科学

IEEE Transactions on Computational Imaging Pub Date : 2024-11-27 DOI: 10.1109/TCI.2024.3507647

Meijun Chen;Zhendong Shi;Wei Chen;Fangjie Xu;Yong Jiang;Yijiang Mao;Shiyue Xu;Bowen Chen;Yalan Wang;Zecheng Wang;Jie Leng

{"title":"An Adaptive Photon Count Reconstruction Algorithm for Sparse Count and Strong Noise Count Data With Low Signal Background Ratio","authors":"Meijun Chen;Zhendong Shi;Wei Chen;Fangjie Xu;Yong Jiang;Yijiang Mao;Shiyue Xu;Bowen Chen;Yalan Wang;Zecheng Wang;Jie Leng","doi":"10.1109/TCI.2024.3507647","DOIUrl":"https://doi.org/10.1109/TCI.2024.3507647","url":null,"abstract":"Single-photon lidar detection data in applications can show different characteristics: sparse count data and strong noise count data with low signal-to-background ratio (SBR), making it difficult to accurately reconstruct depth and intensity information. The existing statistical-based algorithms can achieve reconstruction, but they may lack compatibility for sparse counting and strong noise counting cases which will switch to each other in practical applications. In this paper, an adaptive photon count reconstruction algorithm for sparse count and strong noise count data with low SBR is proposed based on the difference in temporal distribution characteristics between the echo and noise count data. The aggregation characteristic of echo count data in time dimension is proposed to adaptively separate the echo and noise regions in the histogram to reduce the noise interference, and based on the relative difference between count levels in the time neighborhood, an objective function is constructed to reconstruct depth and intensity using optimization. The reconstruction results based on simulated and experimental data confirm that the reconstruction accuracies under both sparse counting and strong noise counting cases are effectively improved under low SBR conditions. Compared with the state-of-the-art algorithms, the depth absolute error is reduced by nearly 50%, the edge error is reduced by an order of magnitude and the proportion of correctly reconstructed pixels reaches 90% when SBR = 0.1. It shows the potential of the proposed algorithm for improving target recognition ability and all-day imaging.","PeriodicalId":56022,"journal":{"name":"IEEE Transactions on Computational Imaging","volume":"10 ","pages":"1799-1814"},"PeriodicalIF":4.2,"publicationDate":"2024-11-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10768988","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142843065","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Learning Light Field Denoising With Symmetrical Refocusing Strategy 利用对称重聚焦策略学习光场去噪

IF 4.2 2区计算机科学

IEEE Transactions on Computational Imaging Pub Date : 2024-11-27 DOI: 10.1109/TCI.2024.3507642

Song Chang;Youfang Lin;Wenqi Wang;Da An;Shuo Zhang

{"title":"Learning Light Field Denoising With Symmetrical Refocusing Strategy","authors":"Song Chang;Youfang Lin;Wenqi Wang;Da An;Shuo Zhang","doi":"10.1109/TCI.2024.3507642","DOIUrl":"https://doi.org/10.1109/TCI.2024.3507642","url":null,"abstract":"Due to hardware restrictions, Light Field (LF) images are often captured with heavy noise, which seriously obstructs the subsequent LF applications. In this paper, we propose a novel symmetrical refocusing strategy to construct the focal stack for every view in LF images and design a simple learning-based framework for LF denoising. Specifically, we first select views that are symmetrically arranged around a target view in LF images. Then we shift and average the selected views to calculate the focal stack, in which all refocused images are aligned with the target view and the noises are effectively suppressed. Then, a Fusion Network is designed to fuse the sharp regions in the focal stack to obtain the denoised target view with sharp details. We further exploit more angular and spatial detail information in LF images and combine the fusion outputs to obtain the final denoised LF images. We evaluate our method in various noise levels and kinds of noisy LF images with different disparity ranges. The experiments show that our method achieves the highest quality in both qualitative and quantitative evaluation than state-of-the-art methods. The proposed symmetrical refocusing strategy is also verified to highly improve the denoising performances.","PeriodicalId":56022,"journal":{"name":"IEEE Transactions on Computational Imaging","volume":"10 ","pages":"1786-1798"},"PeriodicalIF":4.2,"publicationDate":"2024-11-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142821174","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Full-Wave Simulations of Tomographic Optical Imaging Through Augmented Partial Factorization 增广部分分解的层析光学成像全波模拟

IF 4.2 2区计算机科学

IEEE Transactions on Computational Imaging Pub Date : 2024-11-22 DOI: 10.1109/TCI.2024.3499747

Zeyu Wang;Yiwen Zhang;Chia Wei Hsu

{"title":"Full-Wave Simulations of Tomographic Optical Imaging Through Augmented Partial Factorization","authors":"Zeyu Wang;Yiwen Zhang;Chia Wei Hsu","doi":"10.1109/TCI.2024.3499747","DOIUrl":"https://doi.org/10.1109/TCI.2024.3499747","url":null,"abstract":"Label-free optical imaging inside scattering media is important for many disciplines. One challenge is that the ground-truth structure is often unknown, so one cannot rigorously assess and compare different imaging schemes. Full-wave simulations can address this issue, but the heavy computing cost has restricted them to small, typically weakly scattering, systems. Here we use a recently introduced “augmented partial factorization” method to enable full-wave simulations of tomographic optical imaging deep inside multiple-scattering media. We also provide a unifying framework that models different scattering-based imaging methods including reflectance confocal microscopy, optical coherence tomography and microscopy, interferometric synthetic aperture microscopy, and the recently proposed scattering matrix tomography in the same virtual setup, so they can be directly compared to the ground truth and against each other. The ground truth enables the identification of artifacts that would typically be mistaken as being correct while setting a rigorous and uniform standard across different methods. By leveraging the latest advances in computational electromagnetics, this work brings the power, versatility, and convenience of full-wave modeling to deep imaging in the multiple-scattering regime.","PeriodicalId":56022,"journal":{"name":"IEEE Transactions on Computational Imaging","volume":"10 ","pages":"1775-1785"},"PeriodicalIF":4.2,"publicationDate":"2024-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10764744","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142821175","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Relaxation-Based Super-Resolution Method in Pulsed Magnetic Particle Imaging 脉冲磁粒子成像中基于弛豫的超分辨率方法

IF 4.2 2区计算机科学

IEEE Transactions on Computational Imaging Pub Date : 2024-11-20 DOI: 10.1109/TCI.2024.3503364

Lei Li;Haohao Yan;Yuge Li;Yidong Liao;Yanjun Liu;Ruili Zhang;Zhongliang Wang;Xin Feng;Jie Tian

{"title":"Relaxation-Based Super-Resolution Method in Pulsed Magnetic Particle Imaging","authors":"Lei Li;Haohao Yan;Yuge Li;Yidong Liao;Yanjun Liu;Ruili Zhang;Zhongliang Wang;Xin Feng;Jie Tian","doi":"10.1109/TCI.2024.3503364","DOIUrl":"https://doi.org/10.1109/TCI.2024.3503364","url":null,"abstract":"Spatial resolution is one of the most critical indicators for magnetic particle imaging (MPI). Due to factors such as relaxation effects and suboptimal magnetization response, MPI has not yet reached the promised spatial resolution. Pulsed MPI is a method that enables MPI to achieve the resolution predicted by the Langevin function, which thereby enables larger magnetic particles (MNPs) to enhance resolution. To further exceed this resolution, we propose a relaxation-based super-resolution method which leverages the principle that MNPs at different positions exhibit varying relaxation times due to the different DC fields provided by the gradient field. This principle allows the super-resolution method to extract signals from the center of the field free region (FFR) to enhance spatial resolution. The super-resolution method first truncates the exponential decay signal during the plateau phase of the excitation field. Then, the truncated signals are decomposed based on their relaxation times. Finally, signals from the center position of the FFR are retained, and signals from the periphery of the FFR are discarded. Using this retained signal for reconstruction results in a higher spatial resolution. We validate this method via both simulation and experimental measurements. The results indicate that, compared with sinusoidal MPI and pulsed MPI without super-resolution, the super-resolution method has two-fold improvement in resolution.","PeriodicalId":56022,"journal":{"name":"IEEE Transactions on Computational Imaging","volume":"10 ","pages":"1692-1705"},"PeriodicalIF":4.2,"publicationDate":"2024-11-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142789117","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

An Efficient Method for Modelling Millimeter-Wave Scan for Security Screening of Humans 用于人体安全检查的毫米波扫描建模的高效方法

IF 4.2 2区计算机科学

IEEE Transactions on Computational Imaging Pub Date : 2024-11-15 DOI: 10.1109/TCI.2024.3487393

Wenyi Shao;Yan Li

引用次数: 0

Multi-Layered Surface Estimation for Low-Cost Optical Coherence Tomography 低成本光学相干层析成像的多层表面估计

IF 4.2 2区计算机科学

IEEE Transactions on Computational Imaging Pub Date : 2024-11-13 DOI: 10.1109/TCI.2024.3497602

Joshua Rapp;Hassan Mansour;Petros Boufounos;Toshiaki Koike-Akino;Kieran Parsons

{"title":"Multi-Layered Surface Estimation for Low-Cost Optical Coherence Tomography","authors":"Joshua Rapp;Hassan Mansour;Petros Boufounos;Toshiaki Koike-Akino;Kieran Parsons","doi":"10.1109/TCI.2024.3497602","DOIUrl":"https://doi.org/10.1109/TCI.2024.3497602","url":null,"abstract":"Optical coherence tomography (OCT) has broad applicability for 3D sensing, such as reconstructing the surface profiles of multi-layered samples in industrial settings. However, accurately determining the number of layers and their precise locations is a challenging task, especially for low-cost OCT systems having low signal-to-noise ratio (SNR). This paper introduces a principled and noise-robust method of detection and estimation of surfaces measured with OCT. We first derive the maximum likelihood estimator (MLE) for the position and reflectivity of a single opaque surface. We next derive a threshold that uses the acquisition noise variance and the number of measurements available to set a target probability for false acceptance of spurious surface estimates. The threshold and MLE are then incorporated into an algorithm that sequentially detects and estimates surface locations. We demonstrate reconstruction of fine details in samples with optical path lengths around 1 mm and depth error down to 1.5 \u0000<inline-formula><tex-math>$mathrm{mu }$</tex-math></inline-formula>\u0000m despite SNRs as low as –10 dB.","PeriodicalId":56022,"journal":{"name":"IEEE Transactions on Computational Imaging","volume":"10 ","pages":"1706-1721"},"PeriodicalIF":4.2,"publicationDate":"2024-11-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142789047","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0