检索结果-内蒙古大学图书馆

IEEE Aerospace conference

作者： Rojas-Rodriguez, Leandro Rodriguez-Ferreira, Julian Salgado-Meza, Pedro A. Rincon, Sonia Acero, Ignacio F. Hernandez, Francisco L. Palencia-Solano, Francisco Arciniegas-Arbelaez, Daniel Univ Ind Santander E3T Bucaramanga 680002 Colombia Fuerza Aeroespacial Colombiana CITAE Cali 760011 Colombia USA Bogota 111711 Colombia Univ Valle EIEE Cali 760042 Colombia

ISBN: (纸本)9798350304626

Colombia is an emerging space nation currently transitioning from being an operator of space systems to becoming a satellite manufacturer. The LEOPAR mission represents Colombia's fourth satellite development endeavor. It entails a 3U CubeSat platform equipped with a hyperspectral camera known as ANFA, designed for remote sensing in the spectral range of 450 to 900 nm. Its primary objective is to identify vegetation and deforested areas across Colombian territory. The design and implementation of ANFA, in its initial phase as a laboratory prototype, underwent multiple iterations to align with mission requirements based on the state of the art. Rigorous laboratory tests have successfully validated the payload's proper operation in terms of optical, mechanical, and electronic aspects of operational concepts. The ANFA laboratory prototype was developed using Commercial Off-The-Shelf (COTS) devices to emulate the optical subsystem and integrate it with the electronic subsystem. This prototype achieved scene capture, data transmission through the SPI protocol to the instrument's main microcontroller, data processing, storage in an SD memory card, and image reconstruction to identify spectral signatures across each pixel in the hyperspectral data cube. Simultaneously, significant progress has been made in designing the optoelectronic detection chain for ANFA, based on the development of a radiometric model for analyzing a reference scene and transforming scene reflectance to Top of the Atmosphere (TOA). Subsequently, we plan to calculate the Signal-to-Noise Ratio (SNR) performance parameter as a tool to select a suitable detector for our hyperspectral camera, meeting the requirements of the LEOPAR mission. Once this process is completed, the goal is to scale the prototype to an engineering model and, ultimately, a flight model. The vision is for ANFA to become the first Colombian hyperspectral instrument launched into space, thereby aligning this development with ongoing

关键词： image reconstruction

来源：评论

学校读者我要写书评

暂无评论

SNeS: Learning Probably Symmetric Neural Surfaces from incomplete data 17th

SNeS: Learning Probably Symmetric Neural Surfaces from Incom...

引用

17th European conference on Computer Vision (ECCV)

作者： Insafutdinov, Eldar Campbell, Dylan Henriques, Joao F. Vedaldi, Andrea Univ Oxford Oxford OX1 3PJ England

ISBN: (纸本)9783031198236;9783031198243

We present a method for the accurate 3D reconstruction of partly-symmetric objects. We build on the strengths of recent advances in neural reconstruction and rendering such as Neural Radiance Fields (NeRF). A major shortcoming of such approaches is that they fail to reconstruct any part of the object which is not clearly visible in the training image, which is often the case for in-the-wild images and videos. When evidence is lacking, structural priors such as symmetry can be used to complete the missing information. However, exploiting such priors in neural rendering is highly non-trivial: while geometry and non-reflective materials may be symmetric, shadows and reflections from the ambient scene are not symmetric in general. To address this, we apply a soft symmetry constraint to the 3D geometry and material properties, having factored appearance into lighting, albedo colour and reflectivity. We evaluate our method on the recently introduced CO3D dataset, focusing on the car category due to the challenge of reconstructing highly-reflective materials. We show that it can reconstruct unobserved regions with high fidelity and render high-quality novel view images.

关键词： 3D reconstruction Novel view synthesis Neural rendering

来源：评论

学校读者我要写书评

暂无评论

Deep learning filtered back projection for photoacoustic tomography

Deep learning filtered back projection for photoacoustic tom...

引用

Photons Plus Ultrasound: Imaging and Sensing 2025

作者： Tian, Chao Sun, Junjun Liu, Songde University of Science and Technology of China Anhui Hefei230026 China Institute of Artificial Intelligence Hefei Comprehensive National Science Center Anhui Hefei230088 China School of Artificial Intelligence Anhui University Anhui Hefei230601 China

ISBN: (纸本)9781510683860

image reconstruction in photoacoustic tomography (PAT) faces significant challenges due to incomplete projection data. To address this problem, we propose a physics-driven, deep learning-based filtered back-projection algorithm (dFBP). The dFBP framework is systematically designed based on the principles of the traditional analytical filtered back-projection algorithm and consists of a filtering module, a back-projection module, and a fusion module arranged in a cascaded architecture. Numerical simulations show that dFBP-based PAT accurately reconstructs high-quality, artifact-suppressed images from sparse, limited, and acoustically heterogeneous projection data. © 2025 SPIE.

关键词： image restoration reconstruction algorithms Deep learning image quality Photoacoustic tomography

来源：评论

学校读者我要写书评

暂无评论

Intensity-image reconstruction Using Event Camera data by Changing in LSTM Update 13

Intensity-Image Reconstruction Using Event Camera Data by Ch...

引用

13th International conference on Computer and Knowledge Engineering, ICCKE 2023

作者： Soltangholi, Arezoo Rahmati Harati, Ahad Vahedian, Abedin Ferdowsi University of Mashhad Dep. of Computer Engineering Mashhad Iran

ISBN: (纸本)9798350330151

Event cameras offer many advantages, but their output is inherently ambiguous and needs to be converted into a more understandable output. One way to use the output of these cameras is to reconstruct the intensity. Various methods have been proposed for image reconstruction using event data, each attempting to improve image quality from specific aspects. In this study, we aim to increase image quality in a challenging condition when the number of events is very low or zero without retraining or changing the network structure during training. Another challenging situation is at the initial start-up moment which requires an initialization time. In this study, we used the potential of the E2VID model and increased the video quality without changing the trained model. Our method performs better than the E2VID method with an 11.9% improvement in the first 10 frames and a 2% improvement in entire videos in SSIM metric © 2023 IEEE.

关键词： image reconstruction

来源：评论

学校读者我要写书评

暂无评论

Color Crop image Compression Technique using Singular Vector Sparse reconstruction 7

Color Crop Image Compression Technique using Singular Vector...

引用

7th IEEE conference on Information and Communication Technology, CICT 2023

作者： Mishra, Deepak Kumar, Anil Rathor, Vijaypal Singh Pal, Hardev Singh Singh, G.K. PDPM-IIITDM Jabalpur Jabalpur India Iit Dept. of Ee Uttarakhand Roorkee India

ISBN: (纸本)9798350305173

In recent times, applications of crop image analysis is continuously increasing in agriculture fields for crop health monitoring, disease identification, and prevention. Crop images are generally captured by wireless sensor devices using High-resolution cameras, which causes difficulties in the transmission and storage of image data. In this regard, this paper presented singular vector sparse reconstruction (SVSR) method proposed by Xu et al. The main concept of this technique is based on the selection of coefficients and discarding the singular vector obtained from singular value decomposition (SVD) to improve the compression ratio. It is efficient in reconstructing a low-rank singular vector from very few observed entries. In this way, a low-rank approximation of the image data matrix is achieved by SVD, followed by sparse sampling of singular vectors up to a desired level. The reconstruction is performed by using the Makima interpolation (MI) on the sparse samples selected singular vector. This technique's performance has been examined with MI, on color crop image data, in terms of following parameters peak signal-to-noise ratio (PSNR), structural similarity index (SSIM) and compression ratio (CR) as well as visual appearance. According to experimental results, MI interpolation with the SVSR technique achieved 90-98% of compression with good quality of the compressed image. © 2023 IEEE.

关键词： Crops

来源：评论

学校读者我要写书评

暂无评论

Mask Matching Network for Self-supervised Few-shot Medical image Segmentation

Mask Matching Network for Self-supervised Few-shot Medical I...

引用

IEEE International conference on Multimedia and Expo (ICME)

作者： Zhao, Zeyun Wang, Rong Gao, Jianzhe Luo, Zhiming Li, Shaozi Xiamen Univ Dept Artificial Intelligence Xiamen Peoples R China Xiamen Univ Inst Artificial Intelligence Xiamen Peoples R China Wuyi Univ Fujian Key Lab Big Data Applicat & Intellectualiz Jiangmen Peoples R China

ISBN: (纸本)9798350390155;9798350390162

Existing few-shot segmentation methods have achieved remarkable progress in medical image segmentation. However, many existing methods yield incomplete and discontinuous boundary predictions. In contrast, the Segment Anything Model (SAM) consistently produces clear, continuous, and comprehensive segmentation boundaries. Building on this observation, we propose a new two-step network called Mask Matching Network (MMNet) to introduce extra knowledge learned by SAM in natural images for few-shot medical image segmentation. Firstly, Q-Net has been utilized to locate some Regions of Interest (RoI) as prompts for SAM, allowing for the automatic generation of masks without relying on manual prompts. Secondly, we propose a novel Mask Matching Module (MMM), which considers both feature similarity and volume similarity as guidance to collaboratively mine the final segmentation from proposal masks. MMNet achieves state-of-the-art performance with remarkable improvements on two widely used datasets, abdominal MR (ABD) and cardiac MR (CMR), under two different settings.

关键词： Medical image Segmentation Few-shot Segmentation Prototype Networks Foundation Model Mask Matching Module

来源：评论

学校读者我要写书评

暂无评论

MACE CT reconstruction for Modular Material Decomposition from Energy Resolving Photon-Counting data

MACE CT Reconstruction for Modular Material Decomposition fr...

引用

conference on Medical Imaging - Physics of Medical Imaging

作者： Jadue, Natalie M. Nagare, Madhuri Maltz, Jonathan S. Buzzard, Gregery T. Bouman, Charles A. Purdue Univ W Lafayette IN 47907 USA GE HealthCare Waukesha WI 53188 USA

ISBN: (纸本)9781510671553;9781510671546

X-ray computed tomography (CT) based on photon counting detectors (PCD) extends standard CT by counting detected photons in multiple energy bins. PCD data can be used to increase the contrast-to-noise ratio (CNR), increase spatial resolution, reduce radiation dose, reduce injected contrast dose, and compute a material decomposition using a specified set of basis materials.(1) Current commercial and prototype clinical photon counting CT systems utilize PCD-CT reconstruction methods that either reconstruct from each spectral bin separately, or first create an estimate of a material sinogram using a specified set of basis materials and then reconstruct from these material sinograms. However, existing methods are not able to utilize simultaneously and in a modular fashion both the measured spectral information and advanced prior models in order to produce a material decomposition. We describe an efficient, modular framework for PCD-based CT reconstruction and material decomposition using on Multi-Agent Consensus Equilibrium (MACE). Our method employs a detector proximal map or agent that uses PCD measurements to update an estimate of the pathlength sinogram. We also create a prior agent in the form of a sinogram denoiser that enforces both physical and empirical knowledge about the material-decomposed sinogram. The sinogram reconstruction is computed using the MACE algorithm, which finds an equilibrium solution between the two agents, and the final image is reconstructed from the estimated sinogram. Importantly, the modularity of our method allows the two agents to be designed, implemented, and optimized independently. Our results on simulated data show a substantial (450%) CNR boost vs conventional maximum likelihood reconstruction when applied to a phantom used to evaluate low contrast detectability.

关键词： computed tomography CT photon counting detectors material decomposition

来源：评论

学校读者我要写书评

暂无评论

Implementation of Deep image Compressors for Raw image reconstruction 7

Implementation of Deep Image Compressors for Raw Image Recon...

引用

7th International conference on Design Innovation for 3 Cs Compute Communicate Control, ICDI3C 2023

作者： Sarkar, Sayantam Sangeetha, T.R. Halaki, Soujanya Mahadeva Srilakshmi, R. Sushma, K.S. Mvj College of Engineering Department of Ece Bengaluru India

ISBN: (纸本)9798350395556

Deep image compressors are extensively practiced presently conducive to reduce the complications of conventional image compression algorithms, namely JPEG, run-Iength-encoding etc. The primary focal point using the image compression algorithm and deep image compressors is to reduce the fidelity errors in the nonlinear standard RGB (SRGB) color space. Anyhow, to train computers to interpret the visual world, that is for computer vision tasks it requires minimally processed raw image data from the image sensors. To reconstruct the original raw-RGB, small amount of metadata are fixed inside the JPEG image. The traditional JPEG encryption is of raw-RGB reconstruction. To accomplish our objective, extensive structure was used that enables deep networks aiming compression to contemplate the raw reconstruction loss and image fidelity errors. To represents this accession in two structures, Network trains sRGB fidelity loss to modify raw reconstruction loss. The integrated approach leads to considerable advancement in PSNR measured by SSIM. © 2023 IEEE.

关键词： image reconstruction

来源：评论

学校读者我要写书评

暂无评论

Iterative excitation with noise rejection techniques for X-ray computed tomography of hollow turbine blades

引用

NONDESTRUCTIVE TESTING AND EVALUATION 2023年第1期38卷 172-188页

作者： Yang, Fuqiang Zhang, Dinghua Huang, Kuidong Yang, Yao Northwestern Polytech Univ Minist Ind & Informat Technol Key Lab High Performance Mfg Aero Engine Xian Peoples R China Minist Educ Engn Res Ctr Adv Mfg Technol Aero Engine Xian Peoples R China

The features of aero-engine hollow turbine blades show a complicated inner air tract, which have a significant influence on the engine's performance. Inspection of the internal structure and flaws of the blades become indispensable. Non-destructive testing, such as computed tomography (CT), is an effective method for detecting internal problems. The purpose of this study is to demonstrate how an iterative excitation can be used to recover an incomplete projection in CT for a turbine blade. Firstly, the variance of the background was gathered as previous information. Then, to make up for the missed sample at the ill-angle position, a noise map was added and filtered as prep work for the forward projection. The original projections were retained, and the revised projections were used to extract additional characteristics from the damaged data. Finally, both simulation and actual tests were studied. The Normalised Mean Square Distance (NMSD) of the reconstructed image was reduced by over 20%. The Structural Similarity Index Measure (SSIM) and the Universal Quality Index (UQI) were both enhanced by at least 60% and 41%, respectively. It was demonstrated that the technique can increase the accuracy of reconstruction for a hollow turbine blade.

关键词： X-ray computed tomography industry CT turbine blade non-destructive testing incomplete projection

来源：评论

学校读者我要写书评

暂无评论

STOCHASTIC DEEP image PRIOR FOR MULTISHOT COMPRESSIVE SPECTRAL image FUSION

STOCHASTIC DEEP IMAGE PRIOR FOR MULTISHOT COMPRESSIVE SPECTR...

引用

IEEE International conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Jacome, Roman Monroy, Brayan Bacca, Jorge Arguello, Henry Univ Ind Santander Bucaramanga 680002 Colombia

ISBN: (纸本)9798350302615

The Deep image Prior (DIP) technique has been successfully employed in Compressive Spectral Imaging (CSI) as a non-data-driven deep model approach. DIP methodology updates the deep network's weights by minimizing a loss function that considers the difference between the measurements and the forward operator of the network's output. However, this method often yields local minima as all the measurements are evaluated at each iteration. This paper proposes a stochastic deep image prior (SDIP) approach, which stochastically trains DIP networks using random subsets of measurements from different CSI sensors in a CSI fusion (CSIF) setting, resulting in the improvement of the convergence through stochastic gradient descent optimization. The proposed SDIP method improves upon the deterministic DIP and requires less computational time since fewer forward operators are required per iteration. The SPID method provides comparable performance against the state-of-the-art CSIF techniques based on supervised data-driven and unsupervised methods, achieving up to 5 dB in the reconstruction.

关键词： Multishot Acquisition Synthetic Aperture Imaging Stochastic Gradient Descent Deep image Prior

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：