检索结果-内蒙古大学图书馆

2016 International Conference Information Technology and Nanotechnology, ITNT 2016

作者： Kirsh, D.V. Kupriyanov, A.V. Image Processing Systems Institute Branch of the Federal Scientific Research Centre Crystallography and Photonics Russian Academy of Sciences Samara Russia Samara National Research University Samara Russia

The article describes parallel implementations of two parametric identification algorithms. The first algorithm is based on the estimation of Bravais unit cell parameters, and the second one estimates Wigner-Seitz cell volumes. The developed parallel implementations are based on the two-tier model of concurrency. An external tier of concurrency uses Message Passing Interface (MPI) technique to distribute parametric identification tasks between computing nodes. An internal tier uses Open Multi-processing (OpenMP) technique to split single steps of the parametric identification algorithms into independent subtasks. Experimental results on multi-processor/multi-core systems have demonstrated almost linear speedup of the parallel implementations, confirmed the effectiveness of the two-tier concurrency model and proved the stability of the identification approach under lattice distortion.

关键词： Crystal lattices

来源：评论

学校读者我要写书评

暂无评论

Zero Resource Anti-Spoofing Detection for Unit Selection Based Synthetic Speech using image Spectrogram Artifacts

Zero Resource Anti-Spoofing Detection for Unit Selection Bas...

引用

Asia-Pacific Signal and Information processing Association Annual Summit and Conference (APSIPA)

作者： Leow, Su Jun Chng, Eng Siong Lee, Chin-hui Nanyang Technol Univ Singapore Singapore Georgia Inst Technol Atlanta GA 30332 USA

ISBN: (纸本)9789881476821

Synthesized speech poses a serious threat to speaker verification systems, which is aggravated by speech synthesis systems becoming more freely available and easily adaptable to a target speaker. This motivated research into synthetic speech detection to circumvent the threat. Although current algorithms are effective in the detection of HMM-based speech synthesizers, unit selection based speech synthesizers remain a serious threat due to its ability to generate spoofing speech which easily overcame existing detectors. Current error rates for their detection is a lot higher than that obtained for other spoofing methods. This paper proposes a detection algorithm to counter unit selection based synthesis speech. It is free of training and exploits presence of artifacts in image spectrogram to perform detection. To the best of our knowledge, this is the first attempt targeted for unit selection based synthesis speech. Experimental results show the effectiveness of the proposed approach.

关键词： Speech synthesis

来源：评论

学校读者我要写书评

暂无评论

Spatially-variant noise filtering in magnetic resonance imaging: A consensus-based approach

引用

KNOWLEDGE-BASED systems 2016年第Aug.15期106卷 264-273页

作者： Gonzalez-Jaime, Luis Vegas-Sanchez-Ferrero, Gonzalo Kerre, Etienne E. Aja-Fernandez, Santiago Univ Ghent Appl Math & Comp Sci Krijgslaan 281-S9 B-9000 Ghent Belgium Univ Valladolid Lab Proc Imagen ETSI Telecomunicac E-47011 Valladolid Spain Harvard Med Sch Brigham & Womens Hosp ACIL 1249 Boylston St Boston MA 02115 USA Univ Politecn Madrid ETSI Telecomunicac Biomed Image Technol Lab Madrid Spain CIBER BBN Madrid Spain

In order to accelerate the acquisition process in multiple-coil Magnetic Resonance scanners, parallel techniques were developed. These techniques reduce the acquisition time via a sub-sampling of the k-space and a reconstruction process. From a signal and noise perspective, the use of a acceleration techniques modify the structure of the noise within the image. In the most common algorithms, like SENSE, the final magnitude image after the reconstruction is known to follow a Rician distribution for each pixel, just like single coil systems. However, the noise is spatially non-stationary, i.e. the variance of noise becomes x-dependent. This effect can also be found in magnitude images due to other processing inside the scanner. In this work we propose a method to adapt well-known noise filtering techniques initially designed to deal with stationary noise to the case of spatially variant Rician noise. The method copes with inaccurate estimates of variant noise patterns in the image, showing its robustness in realistic cases. The method employs a consensus strategy in conjunction with a set of aggregation functions and a penalty function. Multiple possible outputs are generated for each pixel assuming different unknown input parameters. The consensus approach merges them into a unique filtered image. As a filtering technique, we have selected the Linear Minimum Mean Square Error (LMMSE) estimator for Rician data, which has been used to test our methodology due to its simplicity and robustness. Results with synthetic and in vivo data confirm the good behavior of our approach. (C) 2016 Elsevier B.V. All rights reserved.

关键词： MRI Rician noise SENSE Non-stationary distributed noise Consensus Noise filtering

来源：评论

学校读者我要写书评

暂无评论

A review of computational methods applied for identification and quantification of atherosclerotic plaques in images

引用

EXPERT systems WITH APPLICATIONS 2016年 46卷 1-14页

作者： Jodas, Danilo Samuel Pereira, Aledir Silveira Tavares, Joao Manuel R. S. Minist Educ Brazil CAPES Fdn BR-70040020 Brasilia DF Brazil Univ Estadual Paulista BR-15054000 Sj Do Rio Preto Brazil Univ Porto Fac Engn Inst Ciencia & Inovacao Engn Mecan & Engn Ind P-4200465 Oporto Portugal

Evaluation of the composition of atherosclerotic plaques in images is an important task to determine their pathophysiology. Visual analysis is still as the most basic and often approach to determine the morphology of the atherosclerotic plaques. In addition, computer-aided methods have also been developed for identification of features such as echogenicity, texture and surface in such plaques. In this article, a review of the most important methodologies that have been developed to identify the main components of atherosclerotic plaques in images is presented. Hence, computational algorithms that take into consideration the analysis of the plaques echogenicity, image processing techniques, clustering algorithms and supervised classification used for segmentation, i.e. identification, of the atherosclerotic plaque components in ultrasound, computerized tomography and magnetic resonance images are introduced. The main contribution of this paper is to provide a categorization of the most important studies related to the segmentation of atherosclerotic plaques and its components in images acquired by the most used imaging modalities. In addition, the effectiveness and drawbacks of each methodology as well as future researches concerning the segmentation and classification of the atherosclerotic lesions are also discussed. (C) 2015 Elsevier Ltd. All rights reserved.

关键词： Stroke Medical imaging image analysis image segmentation

来源：评论

学校读者我要写书评

暂无评论

Towards an efficient high-level modeling of heterogeneous image processing systems

Towards an efficient high-level modeling of heterogeneous im...

引用

2016 TMS/DEVS Symposium on Theory of Modeling and Simulation, TMS/DEVS 2016, Part of the 2016 Spring Simulation Multiconference, SpringSim 2016

作者： Deitsch, Anna Schneider, Vitali Kane, Jennifer Dulz, Winfried German, Reinhard Friedrich-Alexander-University of Erlangen-Nuremberg Department of Computer Science 7 Martensstr. 3 ErlangenD-91058 Germany

High-level formal modeling is a beneficial method for assisting system architects in early performance evaluations of desired hardware/software architectures and for coping with the continuously increasing complexity of heterogeneous systems. Modeling languages like UML allow expressive and concise specifications of algorithms, platforms, and infrastructure artifacts for different domains. Techniques are already available for deriving implementations but also executable simulations from UML-based specification models. We use such derived simulations for early investigation and validation of the behavior of the designed system. However, building specification models is often a very complex and time-consuming process, which highly depends on the designer's knowledge, used development process and modeling languages. In this paper, we introduce a holistic top-down modeling methodology developed for enabling an efficient design of heterogeneous image processing architectures using the standardized UML modeling language. Particularly, the suggested methodology focuses on improving user guidance with clear modeling workflows to achieve a less error-prone model design process and to reduce the system development effort. Moreover, we demonstrate how the specific domain knowledge can be captured directly in a UML conform modeling library and that this library enables us to express image processing algorithms in an efficient manner. ©2016 Society for Modeling & Simulation International (SCS).

关键词： Modeling languages

来源：评论

学校读者我要写书评

暂无评论

A Hybrid Algorithm for Dense Stereo Correspondences in Challenging Indoor Scenes

A Hybrid Algorithm for Dense Stereo Correspondences in Chall...

引用

IEEE International Conference on Imaging systems and Techniques (IST) / IEEE International School on Imaging

作者： Georgousis, Stavros Stentoumis, Christos Doulamis, Nikolaos Voulodimos, Athanasios Natl Tech Univ Athens 9 Iroon Polytech Str GR-15773 Athens Greece

ISBN: (纸本)9781509018178

The algorithms for dense correspondences in stereo images are an extensively researched topic, since it is an essential step in a large number of applications. Despite the fact that the first stereo matching algorithms were proposed some decades ago, novel approaches regarding typical, but also cutting-edge applications, are always in demand. Stereo matching is an inverse, ill-posed problem, which usually depends on the application and the scenario. In this contribution, a hybrid approach for stereo matching is proposed, which is based on graph-cuts optimization (global) and cross-based aggregation (local) under a hierarchical scheme. It is shown that the combined effect of a global method in a coarse layer and a local method in finer layers improves the matching results. This hybrid approach exploits the strengths and ameliorates the weaknesses of the individual global and local algorithms. The resulted disparity map is robust without outliers even in untextured areas and at the same time high fidelity details are accurately represented. This hybrid scheme is evaluated on challenging indoor datasets. It is also computationally efficient for applying it on low-processing power applications.

关键词： stereo vision image matching image reconstruction pattern matching

来源：评论

学校读者我要写书评

暂无评论

Real-time Combustion State Identification via image processing: A Dynamic Data-Driven approach

Real-time Combustion State Identification via Image Processi...

引用

American Control Conference (ACC)

作者： Hauser, Michael Li, Yue Li, Jihang Ray, Asok Penn State Univ Mech & Nucl Engn Dept University Pk PA 16802 USA

ISBN: (纸本)9781467386821

This paper proposes a framework for analyzing video of physical processes as a paradigm of dynamic data-driven application systems (DDDAS). The algorithms were tested on a combustion system under fuel lean and ultra-lean conditions. The main challenge here is to develop feature extraction and information compression algorithms with low computational complexity such that they can be applied to real-time analysis of video captured by a high-speed camera. In the proposed method, image frames of the video is compressed into a sequence of image features. then, these image features are mapped to a sequence of symbols by partitioning of the feature space. Finally, a special class of probabilistic finite state automata (PFSA), called D-Markov machines, are constructed from the symbol strings to extract pertinent features representing the embedded dynamic characteristics of the physical process. This paper compares the performance and efficiency of three image feature extraction algorithms: Histogram of Oriented Gradients, Gabor Wavelets, and Fractal Dimension. The k-means clustering algorithm has been used for feature space partitioning. The proposed algorithm has been validated on experimental data in a laboratory environment combustor with a single fuel-injector.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Machine Learning-Based Framework for Multi-Class Diagnosis of Neurodegenerative Diseases: A Study on Parkinson's Disease

引用

IFAC-PapersOnLine 2016年第7期49卷 990-995页

作者： Singh, Gurpreet Vadera, Meet Samavedham, Lakshminarayanan Lim, Erle Chuen-Hian Department of Chemical and Biomolecular Engineering National University of Singapore Singapore Department of Mechanical Engineering Indian Institute of Technology Gandhinagar India Division of Neurology National University Health System National University of Singapore Singapore

A new era of intelligent medical diagnostics is emerging with the development of machine learning-based algorithms to diagnose neurodegenerative diseases (NDDs). In the present work, we discuss an innovative framework that uses principal component analysis (PCA) for feature extraction, Fisher discriminant ratio (FDR) for feature selection and support vector machines (SVM) for classification of Healthy controls, Parkinson's Disease and SWEDD subjects. We have extended our framework to handle the challenge of multi-class disease diagnosis, wherein, accuracy up to 100% has been achieved. This demonstrates the potential of the present methodology to be developed into a clinical relevant diagnostic and decision support system. © 2016

关键词： Neurodegenerative diseases Artificial intelligence Classification (of information) Computer aided diagnosis Computer aided instruction Decision support systems Feature extraction image processing Knowledge based systems Learning systems Principal component analysis Support vector machines Disease diagnosis Fisher discriminant Healthy controls Medical diagnostics Parkinson's disease

来源：评论

学校读者我要写书评

暂无评论

Robotic Imitation of Human Hand Using RGB Color Gradient Mapping

Robotic Imitation of Human Hand Using RGB Color Gradient Map...

引用

International Conference on Artificial Intelligence and Evolutionary Computations in Engineering systems (ICAIECES)

作者： Thyagaraj, T. Abhishek, Kumar Brunda, N. S. Ranjan, Rakesh Kini, Nikita VTU AICTE Dept Elect & Commun Engn BMSIT VTU Campus Belagavi Belagavi 590018 Karnataka India

ISBN: (纸本)9788132226567;9788132226543

The state-of-the-art developments in robotics reveal a scenario wherein human motion imitation has been developed through advancements in features and performance of sensors. Most of such robots uses flex sensors to acquire the information about the orientation of the body being imitated. However, these are slow and are highly vulnerable to damages and physical changes. There are image processing algorithms that are focused on a robots learning through ANN and are not into imitation without memory. This paper presents an overview of research attempts to use a robotic assembly to imitate a real object using image processing. The paper describes results in force, tactile and visual sensing, sensor-based control, and the configuration of a test-based robot integrated system for imitation of a human hand, using vision through image processing using RGB scheme;and one-on-one mapping for the acquisition of the current position of a body and orientation of the various parts of the robot in same fashion. This research has resulted in the development of a prototype capable of imitation and its applications are in the areas of medicine, defense, safety, and explorations for both research and commercial applications.

关键词： Robotic vision Color mapping Median filtering Robot-arm Embedded module image processing Robotics Robotic arm Human imitation Motion tracking Motion sensing Robotic vision Sensor-based robots Flex sensors Embedded systems Integrated systems Artificial neural networks Matlab RGB scheme

来源：评论

学校读者我要写书评

暂无评论

image processing pipeline model integrating steganographic algorithms for mobile cameras

Image processing pipeline model integrating steganographic a...

引用

2016 ACM International Workshop on Traffic Measurements for Cybersecurity, WTMC 2016

作者： Dahal, Prabhat Peng, Dongming Sharif, Hamid Department of Electrical and Computer Engineering University of Nebraska Lincoln United States

ISBN: (纸本)9781450342841

Mobile phones these days are extensively used to capture and share multimedia right from the device without the basic security processing like watermarking. This necessitates the need for multimedia to be watermarked right within the mobile hardware for copyright protection and authentication. In this paper, we propose an image processing pipelining model for mobile cameras that integrates watermarking algorithm with the image capturing processes in cameras. Differently from existing works reported in literature, the unique contribution in this paper is to have image watermarked while in the early capturing process. In other words, the processes for digital image capturing and watermarking have become as a single non-separable procedure in the camera hardware systems. The paper proposes watermarking raw image data before it becomes display ready within the image capturing process. Also, the watermarking algorithm used to develop the model, without losing generality, caters to the computational efficiency requirement for mobile devices while being robust against image coding at the same time. Simulations results demonstrate that the proposed approach is robust against JPEG compression. © 2016 ACM.

关键词： Computational efficiency

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：