IEEE Transactions on Computational Imaging最新文献

{"title":"Deep Motion Regularizer for Video Snapshot Compressive Imaging","authors":"Zan Chen;Ran Li;Yongqiang Li;Yuanjing Feng;Xingsong Hou;Xueming Qian","doi":"10.1109/TCI.2024.3477262","DOIUrl":"https://doi.org/10.1109/TCI.2024.3477262","url":null,"abstract":"Video snapshot compressive imaging (SCI) samples 3D high-speed video frames with temporally varying spatial modulation and compresses them into a single 2D measurement, and the SCI reconstruction algorithm aims to recover the original high-speed frames from the measurement. However, conventional video SCI systems encounter challenges when raising the frame rate of the reconstructed video. To achieve higher frame rates, the modulation mask's rate must be increased, which in turn leads to an increase in the associated hardware expenses. In this paper, we propose a deep unfolding-based reconstruction framework with optical flow for video SCI. The framework recovers both observed and unobserved frames from measurements, resulting in increased video frame rate. To estimate the optical flow, we transform the video features of the network into optical flow features during the iteration. Then, we design a deep denoiser and an optical flow-based motion regularizer combined with the voxels of coarse reconstructed frames to update the observed and unobserved frames. To improve the performance, we employ group convolution in the network and fuse the optical flow information from different phases to reduce the information loss. We further extend the proposed deep unfolding framework to the reconstruction of color SCI videos. Extensive experiments on benchmark data and real data prove that our proposed method has state-of-the-art reconstruction performance and visual effects.","PeriodicalId":56022,"journal":{"name":"IEEE Transactions on Computational Imaging","volume":"10 ","pages":"1519-1532"},"PeriodicalIF":4.2,"publicationDate":"2024-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142524174","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}

{"title":"Provable Preconditioned Plug-and-Play Approach for Compressed Sensing MRI Reconstruction","authors":"Tao Hong;Xiaojian Xu;Jason Hu;Jeffrey A. Fessler","doi":"10.1109/TCI.2024.3477329","DOIUrl":"https://doi.org/10.1109/TCI.2024.3477329","url":null,"abstract":"Model-based methods play a key role in the reconstruction of compressed sensing (CS) MRI. Finding an effective prior to describe the statistical distribution of the image family of interest is crucial for model-based methods. Plug-and-play (PnP) is a general framework that uses denoising algorithms as the prior or regularizer. Recent work showed that PnP methods with denoisers based on pretrained convolutional neural networks outperform other classical regularizers in CS MRI reconstruction. However, the numerical solvers for PnP can be slow for CS MRI reconstruction. This paper proposes a preconditioned PnP (\u0000<inline-formula><tex-math>$text{P}^{2}$</tex-math></inline-formula>\u0000nP) method to accelerate the convergence speed. Moreover, we provide proofs of the fixed-point convergence of the \u0000<inline-formula><tex-math>$text{P}^{2}$</tex-math></inline-formula>\u0000nP iterates. Numerical experiments on CS MRI reconstruction with non-Cartesian sampling trajectories illustrate the effectiveness and efficiency of the \u0000<inline-formula><tex-math>$text{P}^{2}$</tex-math></inline-formula>\u0000nP approach.","PeriodicalId":56022,"journal":{"name":"IEEE Transactions on Computational Imaging","volume":"10 ","pages":"1476-1488"},"PeriodicalIF":4.2,"publicationDate":"2024-10-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142452742","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}

{"title":"A Quantum Denoising-Based Resolution Enhancement Framework for 250-MHz and 500-MHz Quantitative Acoustic Microscopy","authors":"Sayantan Dutta;Jonathan Mamou","doi":"10.1109/TCI.2024.3473312","DOIUrl":"https://doi.org/10.1109/TCI.2024.3473312","url":null,"abstract":"Quantitative acoustic microscopy (QAM) forms two-dimensional (2D) quantitative maps of acoustic properties of thin tissue sections at a microscopic scale (\u0000<inline-formula><tex-math>$&lt; 8; mu$</tex-math></inline-formula>\u0000m) using very-high-frequency (i.e., \u0000<inline-formula><tex-math>$&gt;$</tex-math></inline-formula>\u0000 200 MHz) ultrasonic excitation. Our custom-made QAM systems employ a 250-MHz or a 500-MHz single-element transducer to produce 2D maps with theoretical spatial resolutions smaller than 8 \u0000<inline-formula><tex-math>$mu$</tex-math></inline-formula>\u0000m and 4 \u0000<inline-formula><tex-math>$mu$</tex-math></inline-formula>\u0000m, respectively. Even with the utilization of these state-of-the-art QAM instruments, spatial resolution still proves insufficient for certain clinical studies. However, designing a QAM system yielding finer resolution (i.e., using a higher-frequency transducer) is expensive and requires expert users. This work proposes a scheme to enhance the spatial resolution of the 2D QAM maps by exploiting an off-the-shelf quantum-based adaptive denoiser (DeQuIP), leveraging the principles of quantum many-body theory. Drawing upon the recent advancement in regularization-by-denoising (RED) for image restoration, we impose this external DeQuIP denoiser as a RED-prior coupled with an analytical solution to address the degradation operators in solving the QAM super-resolution problem. The efficiency of our proposed scheme is demonstrated by improving the resolution of experimental 2D acoustic-impedance maps (2DZMs) generated from data acquired using the 250-MHz and 500-MHz QAM systems. Our scheme demonstrates superior performance in recovering finer and subtle details with enhanced spatial resolution when applied to 2DZMs. For example, a spatial resolution improvement of 40% was achieved when applied to 2DZMs at 250-MHz, outperforming two other state-of-the-art methods, which only yielded 23–32% improvement. These observations highlight the efficacy of the proposed RED scheme.","PeriodicalId":56022,"journal":{"name":"IEEE Transactions on Computational Imaging","volume":"10 ","pages":"1489-1504"},"PeriodicalIF":4.2,"publicationDate":"2024-10-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142518159","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}

{"title":"TMSST-RID: A Target-Oriented SAR-ISAR Imaging Method Based on Synchrosqueezing","authors":"Zhifeng Xie;Lei Cui;Qingyuan Shen;Xiaoqing Wang;Peiqing Yang;Xingyi Su;Haifeng Huang","doi":"10.1109/TCI.2024.3468032","DOIUrl":"https://doi.org/10.1109/TCI.2024.3468032","url":null,"abstract":"Synthetic aperture radar (SAR) is an important high-resolution radar system for target surveillance. However, target motion causes a time-varying Doppler frequency in echo and defocused SAR images. To obtain a focused target image, the inverse SAR synthetic aperture radar (ISAR) algorithm is applied: the image is inverted back to the echo and focused using the ISAR imaging algorithm. Clutter significantly impacts the target focusing effect during the SAR-ISAR imaging process. To improve the SAR-ISAR imaging quality of ship targets in the presence of severe sea clutter interference, this article proposes a range instantaneous Doppler method based on the low-rank sparse decomposition and target-oriented multisynchrosqueezing transform (TMSST-RID), which can enhance the target while suppressing clutter. The target area is separated from the clutter area using the low-rank sparse decomposition method. Then, in the time–frequency (TF) analysis, our proposed TMSST method utilizes the prior of the target area to improve the signal-to-clutter ratio (SCR) of the SAR/ISAR image. Compared with the traditional TF transformation, the proposed TMSST enhances the target by sharpening the TF target representation while simultaneously avoiding clutter enhancement. The superiority of this SAR/ISAR imaging method is demonstrated in a performance evaluation using simulated and real radar data for ship targets in sea clutter.","PeriodicalId":56022,"journal":{"name":"IEEE Transactions on Computational Imaging","volume":"10 ","pages":"1425-1438"},"PeriodicalIF":4.2,"publicationDate":"2024-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142383484","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}

{"title":"Accelerated Wirtinger Flow With Score-Based Image Priors for Holographic Phase Retrieval in Poisson-Gaussian Noise Conditions","authors":"Zongyu Li;Jason Hu;Xiaojian Xu;Liyue Shen;Jeffrey A. Fessler","doi":"10.1109/TCI.2024.3458418","DOIUrl":"https://doi.org/10.1109/TCI.2024.3458418","url":null,"abstract":"Phase retrieval (PR) is a crucial problem in many imaging applications. This study focuses on holographic phase retrieval in situations where the measurements are degraded by a combination of Poisson and Gaussian noise, as commonly occurs in optical imaging systems. We propose a new algorithm called “AWFS” that uses accelerated Wirtinger flow (AWF) with a learned score function as a generative prior. Specifically, we formulate the PR problem as an optimization problem that incorporates both data fidelity and regularization terms. We calculate the gradient of the log-likelihood function for PR and determine its corresponding Lipschitz constant. Additionally, we introduce a generative prior in our regularization framework by using score matching to capture information about the gradient of image prior distributions. We provide theoretical analysis that establishes a critical-point convergence guarantee for one version of the proposed algorithm. The results of our simulation experiments on three different datasets show the following. 1) By using the PG likelihood model, a practical version of the proposed algorithm improves reconstruction compared to algorithms based solely on Gaussian or Poisson likelihoods. 2) The proposed score-based image prior method leads to better reconstruction quality than a method based on denoising diffusion probabilistic model (DDPM), as well as a plug-and-play alternating direction method of multipliers (PnP-ADMM) and regularization by denoising (RED).","PeriodicalId":56022,"journal":{"name":"IEEE Transactions on Computational Imaging","volume":"10 ","pages":"1384-1399"},"PeriodicalIF":4.2,"publicationDate":"2024-09-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142328360","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}

{"title":"Spectral Memory-Enhanced Network With Local Non-Local and Low-Rank Priors for Hyperspectral Image Compressive Imaging","authors":"Yangke Ying;Jin Wang;Yunhui Shi;Nam Ling;Baocai Yin","doi":"10.1109/TCI.2024.3468615","DOIUrl":"https://doi.org/10.1109/TCI.2024.3468615","url":null,"abstract":"The hyperspectral image (HSI) compressive imaging field has experienced significant progress in recent years, especially with the emergence of deep unfolding networks (DUNs), which have demonstrated remarkable advancements in reconstruction performance. However, these methods still face several challenges. Firstly, HSI data carries crucial prior knowledge in the feature space, and effectively leveraging these priors is essential for achieving high-quality HSI reconstruction. Existing methods either neglect the utilization of prior information or incorporate network modules designed based on prior information in a rudimentary manner, thereby limiting the overall reconstruction potential of these models. Secondly, the transformation between the data and feature domains poses a significant challenge for DUNs, leading to the loss of feature information across different stages. Existing methods fall short in adequately considering spectral characteristics when utilizing inter-stage information, resulting in inefficient transmission of feature information. In this paper, we introduce a novel deep unfolding network architecture that integrates local non-local and low-rank priors with spectral memory enhancement for precise HSI data reconstruction. Specifically, we design innovative modules for local non-local and low-rank priors to enrich the network's feature representation capability, fully exploiting the prior information of HSI data in the feature space. These designs also help the overall framework achieve superior reconstruction results with fewer parameters. Moreover, we extensively consider the spectral correlation characteristics of HSI data and devise a spectral memory enhancement network module to mitigate inter-stage feature information loss. Extensive experiments further demonstrate the superiority of our approach.","PeriodicalId":56022,"journal":{"name":"IEEE Transactions on Computational Imaging","volume":"10 ","pages":"1664-1679"},"PeriodicalIF":4.2,"publicationDate":"2024-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142753892","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}

{"title":"Fast Non-Line-of-Sight Imaging With Hybrid Super-Resolution Network Over 18 m","authors":"Leping Xiao;Jianyu Wang;Yi Wang;Ziyu Zhan;Zuoqiang Shi;Lingyun Qiu;Xing Fu","doi":"10.1109/TCI.2024.3463964","DOIUrl":"https://doi.org/10.1109/TCI.2024.3463964","url":null,"abstract":"Non-line-of-sight (NLOS) imaging technique aims at visualizing hidden objects from light of multiple reflections. For most existing methods, densely raster-scanned transients with long exposure time are routinely used, while approaches employing fewer points are confronted with a trade-off between the computation time and the image quality, both of which hinder the practical implementation of fast NLOS imaging. In this paper, we propose a hybrid super-resolution pipeline for image reconstruction and quality enhancement with only 8×8 scanning points. Besides, we implement a non-coaxial transceiver configuration and illustrate the first auto-calibration method for out-of-lab NLOS configuration, which costs only 40 s and performs well at a distance of 18.69 m. Results on both experimental data and public dataset indicate that the proposed method exhibits strong generalization capabilities, yielding faithful reconstructions with the resolution of 256×256 under different noise models. Furthermore, we demonstrate the importance of matching the noise model with the experimental dataset. We believe our approach shows great promise to NLOS imaging acceleration with lower acquisition, calibration and computation time.","PeriodicalId":56022,"journal":{"name":"IEEE Transactions on Computational Imaging","volume":"10 ","pages":"1439-1448"},"PeriodicalIF":4.2,"publicationDate":"2024-09-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142430859","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}

{"title":"3D Helical CT Reconstruction With a Memory Efficient Learned Primal-Dual Architecture","authors":"Jevgenija Rudzusika;Buda Bajić;Thomas Koehler;Ozan Öktem","doi":"10.1109/TCI.2024.3463485","DOIUrl":"10.1109/TCI.2024.3463485","url":null,"abstract":"Deep learning based computed tomography (CT) reconstruction has demonstrated outstanding performance on simulated 2D low-dose CT data. This applies in particular to domain adapted neural networks, which incorporate a handcrafted physics model for CT imaging. Empirical evidence shows that employing such architectures reduces the demand for training data and improves upon generalization. However, their training requires large computational resources that quickly become prohibitive in 3D helical CT, which is the most common acquisition geometry used for medical imaging. This paper modifies a domain adapted neural network architecture, the Learned Primal-Dual (LPD), so that it can be trained and applied to reconstruction in this setting. The main challenge is to reduce the GPU memory requirements during the training, while keeping the computational time within practical limits. Furthermore, clinical data also comes with other challenges not accounted for in simulations, like errors in flux measurement, resolution mismatch and, most importantly, the absence of the real ground truth. To the best of our knowledge, this work is the first to apply an unrolled deep learning architecture for reconstruction on full-sized clinical data, like those in the Low dose CT image and projection data set (LDCT).","PeriodicalId":56022,"journal":{"name":"IEEE Transactions on Computational Imaging","volume":"10 ","pages":"1414-1424"},"PeriodicalIF":4.2,"publicationDate":"2024-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142249426","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}

{"title":"Progressive Self-Supervised Learning for CASSI Computational Spectral Cameras","authors":"Xiaoyin Mei;Yuqi Li;Qiang Fu;Wolfgang Heidrich","doi":"10.1109/TCI.2024.3463478","DOIUrl":"10.1109/TCI.2024.3463478","url":null,"abstract":"Compressive spectral imaging (CSI) is a technique used to capture high-dimensional hyperspectral images (HSIs) with a few multiplexed measurements, thereby reducing data acquisition costs and complexity. However, existing CSI methods often rely on end-to-end learning from training sets, which may struggle to generalize well to unseen scenes and phenomena. In this paper, we present a progressive self-supervised method specifically tailored for coded aperture snapshot spectral imaging (CASSI). Our proposed method enables HSI reconstruction solely from the measurements, without requiring any ground truth spectral data. To achieve this, we integrate positional encoding and spectral cluster-centroid features within a novel progressive training framework. Additionally, we employ an attention mechanism and a multi-scale architecture to enhance the robustness and accuracy of HSI reconstruction. Through extensive experiments on both synthetic and real datasets, we validate the effectiveness of our method. Our results demonstrate significantly superior performance compared to state-of-the-art self-supervised CASSI methods, while utilizing fewer parameters and consuming less memory. Furthermore, our proposed approach showcases competitive performance in terms of reconstruction quality when compared to state-of-the-art supervised methods.","PeriodicalId":56022,"journal":{"name":"IEEE Transactions on Computational Imaging","volume":"10 ","pages":"1505-1518"},"PeriodicalIF":4.2,"publicationDate":"2024-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142249424","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}

{"title":"DS$^{2}$PN: A Two-Stage Direction-Aware Spectral-Spatial Perceptual Network for Hyperspectral Image Reconstruction","authors":"Tiecheng Song;Zheng Zhang;Kaizhao Zhang;Anyong Qin;Feng Yang;Chenqiang Gao","doi":"10.1109/TCI.2024.3458421","DOIUrl":"10.1109/TCI.2024.3458421","url":null,"abstract":"Coded aperture snapshot spectral imaging (CASSI) systems are designed to modulate and compress 3D hyperspectral images (HSIs) into 2D measurements, which can capture HSIs in dynamic scenes. How to faithfully recover 3D HSIs from 2D measurements becomes one of the challenges. Impressive results have been achieved by deep leaning methods based on convolutional neural networks and transformers, but the directional information is not thoroughly explored to reconstruct HSIs and evaluate the reconstruction quality. In view of this, we propose a two-stage direction-aware spectral-spatial perceptual network (DS\u0000<inline-formula><tex-math>$^{2}$</tex-math></inline-formula>\u0000PN) for HSI reconstruction. In the first stage, we design a frequency-based preliminary reconstruction subnetwork to roughly recover the global spectral-spatial information of HSIs via frequency interactions. In the second stage, we design a multi-directional spectral-spatial refinement subnetwork to recover the details of HSIs via directional attention mechanisms. To train the whole network, we build a pixel-level reconstruction loss for each subnetwork, and a feature-level multi-directional spectral-spatial perceptual loss which is specially tailored to high-dimensional HSIs. Experimental results show that our DS\u0000<inline-formula><tex-math>$^{2}$</tex-math></inline-formula>\u0000PN outperforms state-of-the-art methods in quantitative and qualitative evaluation for both simulation and real HSI reconstruction tasks.","PeriodicalId":56022,"journal":{"name":"IEEE Transactions on Computational Imaging","volume":"10 ","pages":"1346-1356"},"PeriodicalIF":4.2,"publicationDate":"2024-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142177621","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}