检索结果-内蒙古大学图书馆

OpenCL-based optimization methods for utilizing forward DCT and quantization of image compression on a heterogeneous platform

引用

JOURNAL OF REAL-TIME image processing 2016年第2期12卷 219-235页

作者： Alqudami, Nasser Kim, Shin-Dug Yonsei Univ Dept Comp Sci Coll Engn 134 Sinchon Dong Seoul 120749 South Korea

Recent computer systems and handheld devices are equipped with high computing capability, such as general purpose GPUs (GPGPU) and multi-core CPUs. Utilizing such resources for computation has become a general trend, making their availability an important issue for the real-time aspect. Discrete cosine transform (DCT) and quantization are two major operations in image compression standards that require complex computations. In this paper, we develop an efficient parallel implementation of the forward DCT and quantization algorithms for JPEG image compression using Open Computing Language (OpenCL). This OpenCL-based parallel implementation utilizes a multi-core CPU and a GPGPU to perform DCT and quantization computations. We demonstrate the capability of this design via two proposed working scenarios. The proposed approach also applies certain optimization techniques to improve the kernel execution time and data movements. We developed an optimal OpenCL kernel for a particular device using device-based optimization factors, such as thread granularity, work-items mapping, workload allocation, and vector-based memory access. We evaluated the performance in a heterogeneous environment, finding that the proposed parallel implementation was able to speed up the execution time of the DCT and quantization by factors of 7.97 and 8.65, respectively, obtained from 1024 x 1024 and 2084 x 2048 image sizes in 4:4:4 format.

关键词： Forward DCT Quantization image compression Parallel image processing Heterogeneous computing OpenCL

来源：评论

学校读者我要写书评

暂无评论

Digital Signal processing: Principles, algorithms and System Design (Instructor's Edu Resource 1 of 2, Solution Manual & Lectures) (Solutions) 1

引用

2016年

作者： Winser Alexander Cranos M Williams

ISBN: (纸本)0128045477;9780128045473

Digital signal processing (DSP) has been applied to a very wide range of applications. This includes voice processing, image processing, digital communications, the transfer of data over the internet, image and data compression, etc. Engineers who develop DSP applications today, and in the future, will need to address many implementation issues including mapping algorithms to computational structures, computational efficiency, power dissipation, the effects of finite precision arithmetic, throughput and hardware implementation. It is not practical to cover all of these in a single text. However, this text emphasizes the practical implementation of DSP algorithms as well as the fundamental theories and analytical procedures that form the basis for modern DSP applications. Digital Signal processing: Principles, algorithms and System Design provides an introduction to the principals of digital signal processing along with a balanced analytical and practical treatment of algorithms and applications for digital signal processing. It is intended to serve as a suitable text for a one semester junior or senior level undergraduate course. It is also intended for use in a following one semester first-year graduate level course in digital signal processing. It may also be used as a reference by professionals involved in the design of embedded computer systems, application specific integrated circuits or special purpose computer systems for digital signal processing, multimedia, communications, or image processing. Covers fundamental theories and analytical procedures that form the basis of modern DSP Shows practical implementation of DSP in software and hardware Includes Matlab for design and implementation of signal processing algorithms and related discrete time systems Bridges the gap between reference texts and the knowledge needed to implement DSP applications in software or hardware

关键词：

来源：评论

学校读者我要写书评

暂无评论

GPII: A benchmark for generic purpose image information 9

GPII: A benchmark for generic purpose image information

引用

9th International C* Conference on Computer Science and Software Engineering, C3S2E 2016

作者： Martins, Pedro Cecílio, José Abbasi, Maryam Furtado, Pedro University of Coimbra Coimbra Portugal Coimbra Portugal

ISBN: (纸本)9781450340755

The growing number of different models and approaches for Geographic Information systems (GIS) brings high complexity when we want to develop new approaches and compare a new GIS algorithm. In order to test and compare different processing models and approaches, in a simple way, we identified the need of defining uniform testing methods, able to compare processing algorithms in terms of performance and accuracy regarding large image processing, algorithms for GIS pattern-detection. Taking into account, for instance, images collected during a done flight or a satellite, it is important to know the processing cost to extract data when applying different processing models and approaches, as well as their accuracy (compare execution time vs. extracted data quality). In this work, we propose a GIS Benchmark (GPII), a benchmark that allows evaluating different approaches to detect/extract selected features from a GIS dataset. Considering a given dataset (or two data-sets, from different years, of the same region), it provides linear methods to compare different performance parameters regarding GIS information, making possible to access the most relevant information in terms of features and processing efficiency. Moreover, our approach to test algorithms makes possible to change the data-set in order to support different purpose algorithms. © 2016 ACM.

关键词： Geographic information systems

来源：评论

学校读者我要写书评

暂无评论

Euclidean proximity function in image processing

Euclidean proximity function in image processing

引用

2016 Signal processing: algorithms, Architectures, Arrangements, and Applications, SPA 2016

作者： Stasik, Pawel M. Department of Control and Systems Engineering Faculty of Computing Science Poznán University of Technology Poznań Poland

ISBN: (纸本)9788362065271

In image processing measuring and valuing a distance between two points is important. The obtained values can be used for determining whether two points are close to each other or to define weights for a filter concentrated around a central element. While there are measures of proximity, neither of them was defined with a such use in mind, mostly concentrating on problem of an optimization. The idea is to turn the Euclidean distance between two points into a measure of how close (or far) two points are from each other, basing on two given ranges. The function was mostly obtained by a theoretical analysis supported with a mathematical calculation and examples of use. As it was proven in the work, the obtained function can be implemented not only to measure proximity, but also as a flexible kernel for image filters, allowing for blurring or edge-detection.

关键词： Edge detection

来源：评论

学校读者我要写书评

暂无评论

Multimodal Task-Driven Dictionary Learning for image Classification

引用

IEEE TRANSACTIONS ON image processing 2016年第1期25卷 24-38页

作者： Bahrampour, Soheil Nasrabadi, Nasser M. Ray, Asok Jenkins, William Kenneth Penn State Univ Dept Elect Engn University Pk PA 16802 USA Bosch Res & Technol Ctr Palo Alto CA 94304 USA US Army Res Lab Adelphi MD 20783 USA W Virginia Univ Comp Sci & Elect Engn Dept Morgantown WV 26506 USA

Dictionary learning algorithms have been successfully used for both reconstructive and discriminative tasks, where an input signal is represented with a sparse linear combination of dictionary atoms. While these methods are mostly developed for single-modality scenarios, recent studies have demonstrated the advantages of feature-level fusion based on the joint sparse representation of the multimodal inputs. In this paper, we propose a multimodal task-driven dictionary learning algorithm under the joint sparsity constraint (prior) to enforce collaborations among multiple homogeneous/heterogeneous sources of information. In this task-driven formulation, the multimodal dictionaries are learned simultaneously with their corresponding classifiers. The resulting multimodal dictionaries can generate discriminative latent features (sparse codes) from the data that are optimized for a given task such as binary or multiclass classification. Moreover, we present an extension of the proposed formulation using a mixed joint and independent sparsity prior, which facilitates more flexible fusion of the modalities at feature level. The efficacy of the proposed algorithms for multimodal classification is illustrated on four different applications-multimodal face recognition, multi-view face recognition, multi-view action recognition, and multimodal biometric recognition. It is also shown that, compared with the counterpart reconstructive-based dictionary learning algorithms, the task-driven formulations are more computationally efficient in the sense that they can be equipped with more compact dictionaries and still achieve superior performance.

关键词： Dictionary learning multimodal classification sparse representation feature fusion

来源：评论

学校读者我要写书评

暂无评论

High speed Vedic Multiplier for image processing using FPGA 10

High speed Vedic Multiplier for Image processing using FPGA

引用

10th International Conference on Intelligent systems and Control (ISCO)

作者： Jayakumar, S. Sumathi, S. Adhiyamaan Coll Engn Dept EIE Hosur India Adhiyamaan Coll Engn Dept ECE Hosur India

ISBN: (纸本)9781467378079

Vedic multiplier is based on the ancient algorithms (sutras) followed in INDIA for multiplication. This work is based on one of the sutras called "Nikhilam Sutra". This sutra is meant for faster mental calculation. Though faster when implemented in hardware, it consumes more power than the conventional ones. This paper presents a technique to modify the architecture of the Vedic multiplier by using some existing methods in order to reduce power and improve image processing applicatio. The 32 X 32 Vedic multiplier is coded in Verilog HDL and Synthesized using Synopsys Design Compiler. The performance is compared in terms of area, data arrival time and power with earlier existing architecture of Vedic multiplier. Filtering involves lots of multiplications which consumes time. Time required increases with the increase in the number of pixels. This paper proposes an approach for image filtering using Vedic Mathematic which performs faster multiplication compared to the conventional algorithms namely Booth and Array Multiplication Algorithm thus reducing the time required for filtering of images. Time required by the algorithms for filtering are then compared using the experimental results.

关键词： image filtering Nikhilam sutra Vedic multiplier

来源：评论

学校读者我要写书评

暂无评论

Assessing particle kinematics via template matching algorithms

引用

OPTICS EXPRESS 2016年第8期24卷 7987-8012页

作者： Weber, M. Fink, M. Fortov, V. Lipaev, A. Molotkov, V. Morfill, G. Petrov, O. Pustylnik, M. Thoma, M. Thomas, H. Usachev, A. Raeth, C. Deutsch Zentrum Luft & Raumfahrt Forsch Grp Komplexe Plasmen Argelsrieder Feld 1a D-82234 Wessling Germany Univ Munich Geschwister Scholl Pl 1 D-80539 Munich Germany Russian Acad Sci JIHT Joint Inst High Temp Izhorskaya St 13 Bd 2 Moscow 125412 Russia Max Planck Inst Extraterr Phys D-85748 Garching Germany Univ Giessen Goethest 58 D-35390 Giessen Germany

Template matching algorithms represent a viable tool to locate particles in optical images. A crucial factor of the performance of these methods is the choice of the similarity measure. Recently, it was shown in [Gao and Helgeson, Opt. Express 22 (2014)] that the correlation coefficient (CC) leads to good results. Here, we introduce the mutual information (MI) as a nonlinear similarity measure and compare the performance of the MI and the CC for different noise scenarios. It turns out that the mutual information leads to superior results in the case of signal dependent noise. We propose a novel approach to estimate the velocity of particles which is applicable in imaging scenarios where the particles appear elongated due to their movement. By designing a bank of anisotropic templates supposed to fit the elongation of the particles we are able to reliably estimate their velocity and direction of motion out of a single image. (C) 2016 Optical Society of America

关键词： image registration Imaging systems Medical image processing Speckle noise Synthetic aperture radar X ray imaging

来源：评论

学校读者我要写书评

暂无评论

Linear in the parameter model for Homomorphic filter: Volterra series based approach 11

Linear in the parameter model for Homomorphic filter: Volter...

引用

11th International Conference on Industrial and Information systems (ICIIS)

作者： Venkatappareddy, P. Lall, Brejesh Indian Inst Technol Dept Elect Engn New Delhi 110016 India

ISBN: (纸本)9781509038183

This paper introduces a linear in the parameter model for Homomorphic filter using Volterra series approach. To obtain these parameters we propose a model where we choose a sub image from the response of Homomorphic filter as reference image to reduce computational complexity. We apply non uniform illuminated images to the proposed filter and compare its performance against standard Homomorphic filter. The proposed filter outperforms the traditional Homomorphic filter in all experiments. Also we compare the error convergence and steady-state error of Sparse aware LMS with LMS algorithm to calculate proposed filter coefficients.

关键词： Lighting Computational modeling Convergence image processing Mathematical model Kernel Signal processing algorithms

来源：评论

学校读者我要写书评

暂无评论

Developing a web-based system for supervised classification of remote sensing images

引用

GEOINFORMATICA 2016年第4期20卷 629-649页

作者： Sun, Ziheng Fang, Hui Di, Liping Yue, Peng Tan, Xicheng Bai, Yuqi George Mason Univ CSISS 4400 Univ DrMSN 6E1 Fairfax VA 22030 USA Wuhan Univ LIESMARS Wuhan Hubei Peoples R China Wuhan Univ Int Sch Software Spatial Informat & Digital Technol Dept Wuhan Hubei Peoples R China Tsinghua Univ Ctr Earth Syst Sci Minist Educ Key Lab Earth Syst Modeling Beijing 100084 Peoples R China

Web-based image classification systems aim to provide users with an easy access to image classification function. The existing work mainly focuses on web-based unsupervised classification systems. This paper proposes a web-based supervised classification system framework which includes three modules: client, servlet and service. It comprehensively describes how to combine the procedures of supervised classification into the development of a web system. A series of methods are presented to realize the modules respectively. A prototype system of the framework is also implemented and a number of remote sensing (RS) images are tested on it. Experiment results show that the prototype is capable of accomplishing supervised classification of RS images on the Web. If appropriate algorithms and parameter values are used, the results of the web-based solution could be as accurate as the results of traditional desktop-based systems. This paper lays the foundation on both theoretical and practical aspects for the future development of operational web-based supervised classification systems.

关键词： Supervised classification Geoprocessing web service Web-based processing system Remote sensing image Cyberinsfrastructure

来源：评论

学校读者我要写书评

暂无评论

Automatic Thresholding Method for Edge Detection algorithms 8

Automatic Thresholding Method for Edge Detection Algorithms

引用

8th International Conference on Electronics, Computers and Artificial Intelligence (ECAI)

作者： Anghelescu, Petre Iliescu, Vlad Gabriel Mara, Constantin Gavriloaia, Mihai-Bogdan Univ Pitesti Dept Elect Commun & Comp Pitesti Romania Univ Politehn Bucuresti Telecommun Dpt Elect Bucharest Romania

ISBN: (纸本)9781509020478

This paper describes an efficient edge detection algorithm that can be used as a plug-in for digital image processing systems. The proposed algorithm uses a method based on iterative clustering targeting a reduced number of operations. The algorithm splits the image into two parts, background and foreground, and calculates the mean value for each of them. Based on these results, the new threshold value will be obtained and looped until the mean values remain unchanged. The only pixels affected by the change are the pixels with values between the previous two thresholds, so only they have to be redistributed to a new class. As a result, only few operations are needed in order to obtain the desired threshold. All the algorithms and results obtained in this paper are developed and tested using the C# programming language.

关键词： image processing edge detection thresholding algorithms non-maximum suppression cluster based methods

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：