检索结果-内蒙古大学图书馆

IEEE International Conference on Fuzzy Systems (FUZZ-IEEE)

作者： S. Pemmaraju S. Mitra Computer Vision and Image Analysis Laboratory Department of Electrical Engineering Texas Tech University Lubbock TX USA

Since most real data sets encountered in cluster analysis are contaminated with background noise or outliers, it is essential to detect and isolate these noise samples from the data set. A few noisy points can affect the clustering procedure by severely biasing the algorithm. An ideal solution to this problem is to identify all the outliers and form a separate noise cluster. To do so, a technique is required by which the noise points are automatically identified and removed from the pattern data. The authors present a modified adaptive fuzzy leader clustering (AFLC) algorithm that has been used to detect and eliminate the outliers from the data structure and create a separate cluster of the outliers. The AFLC algorithm has an adaptive resonance theory (ART) like architecture with fuzzy learning embedded into it. Test results of the algorithm when applied to real and noisy data sets are presented.< >

关键词： Intelligent networks Fuzzy neural networks Clustering algorithms Background noise Subspace constraints image analysis Data structures Neural networks computer vision Laboratories

来源：评论

学校读者我要写书评

暂无评论

A new method for estimating dimensions and 3-D reconstruction of coronary arterial trees from biplane angiograms

A new method for estimating dimensions and 3-D reconstructio...

引用

Annual IEEE Symposium on computer-Based Medical Systems

作者： T. Kayikcioglu S. Mitra Computer Vision and Image Analysis Laboratory Department of Electrical Engineering Texas Tech University Lubbock TX USA

The coronary arteries may be represented and visualized efficiently by a generalized cylinder (GC) model with elliptical cross-sections. By using this model and utilizing quadratic expression of an ellipse, the intensity distribution of an ellipse in any image plane, which is perpendicular to the plane of the arterial cross-section, is computed. A nonlinear parametric model for observed intensity distribution of an arterial cross-section is applied to estimate not only cross-sectional areas but also parameters of the ellipse which are essential in 3-D reconstruction of an arterial segment. This model takes into account background intensity, noise and blurring. Performance of this model has been tested on computer-generated and actual data for various background types and noise values and compared to those of other methods for estimating arterial dimensions. The computed cross-sectional areas from our model for computer-generated as well as actual data demonstrate much less variability than those reported in recent literature. 3-D reconstruction of arterial segments from actual biplane angiograms is presented.< >

关键词： Three dimensional displays Arteries Background noise Diseases Parameter estimation Visualization Parametric statistics image reconstruction Shape Bones

来源：评论

学校读者我要写书评

暂无评论

Integrated adaptive fuzzy clustering (IAFC) algorithm

Integrated adaptive fuzzy clustering (IAFC) algorithm

引用

IEEE International Conference on Fuzzy Systems (FUZZ-IEEE)

作者： Y.S. Kim S. Mitra Computer Vision and Image Analysis Laboratory Department of Electrical Engineering Texas Tech University Lubbock TX USA

An integrated adaptive fuzzy clustering (IAFC) algorithm using a structure similar to that found in the Adaptive Resonance Theory (ART-1) neural network, is presented. The IAFC incorporates a new learning rule and a new similarity measure to eliminate some structural problems inherent in other fuzzy ART-type neural networks. The new learning rule utilizes a fuzzy membership value, a function of the number of iterations, and a fuzzy within-cluster membership value. The new similarity measure incorporates a fuzzy membership value to the Euclidean distance. This incorporation of the new learning rule and the new similarity measure guarantees the convergence of weights in the IAFC algorithm and provides more flexibility to the shapes of the clusters formed by this algorithm. The critical parameters in the operation of IAFC are discussed. The performance of IAFC is evaluated in the classification of real data and compared with other recent neuro-fuzzy clustering algorithms.< >

关键词： Clustering algorithms Neural networks Fuzzy neural networks Euclidean distance Partitioning algorithms Fuzzy sets Resonance Shape measurement Convergence Subspace constraints

来源：评论

学校读者我要写书评

暂无评论

Comments on "A theoretical analysis of Monte Carlo algorithms for the simulation of Gibbs random field images" [with reply]

引用

IEEE Transactions on Information Theory 1993年第1期39卷 310-311页

作者： R.M. Neal J.K. Goutsias Department of Computer Science University of Toronto Toronto Canada Department of Electrical and Computer Engineering Image Analysis and Communications Laboratory Johns Hopkins University Baltimore MD USA

The commenter contends that, contrary to the conclusions of the above-titled paper by J.K. Goutsias (see ibid., vol.37, no.6, p.1618-28, Nov. 1991), the relative entropy of the initial distribution for a Monte Carlo simulation is not monotonically related to the degree of convergence after some given number of steps, and optimal single-step convergence is not achieved by the Gibbs sampler with systematic site ordering. In his reply, Goutsias explains the intent of his work, but expresses agreement with many of the comments. He provides a corrected statement of Theorem 3.< >

关键词： Algorithm design and analysis Monte Carlo methods Analytical models Convergence Entropy image converters Equations Sampling methods image processing image analysis

来源：评论

学校读者我要写书评

暂无评论

An analysis of Monte Carlo methods for likelihood estimation of Gibbsian images

An analysis of Monte Carlo methods for likelihood estimation...

引用

1993 IEEE International Conference on Acoustics, Speech, and Signal Processing

作者： G. Potamianos J. Goutsias Department of Electrical and Computer Engineering Image Analysis and Communications Laboratory Johns Hopkins University Baltimore MD USA

A unified analysis of two Monte Carlo algorithms for estimating the likelihood function of Gibbs random field images is presented. It is shown that such an estimation reduces to estimating the partition functions of suitably chosen Gibbs random fields. The first algorithm requires drawing samples from a mutually compatible Gibbs random field and provides unbiased and consistent estimators of these partition functions. The second algorithm uses samples which are approximately drawn from the Gibbs distribution, and results in asymptotically unbiased and consistent estimators of the partition functions. A measure of the computational complexity of these algorithms which makes it possible to compare them is introduced. It is concluded that the first algorithm is superior, especially for models with strong interactions.< >

关键词： image analysis Partitioning algorithms Monte Carlo methods Computational complexity Testing Laboratories image processing Parameter estimation Scattering Algorithm design and analysis

来源：评论

学校读者我要写书评

暂无评论

Filtering of uncertain irregularly sampled multidimensional data

Filtering of uncertain irregularly sampled multidimensional ...

引用

Asilomar Conference on Signals, Systems & computers

作者： H. Knutsson C.-F. Westin C.-J. Westelius Computer Vision Laboratory Department of Electrical Engineering Linköping University Linkoping Sweden

Two new methods are presented: normalized convolution and normalized differential convolution. These methods are examples of the power of the signal/certainty-philosophy, i.e. the separation of both data and operator into a signal part and a certainty part. Missing data is simply handled by setting the certainty to zero. Localization or 'windowing' of operators is done using an applicability function, the operator equivalent to certainty, not by changing the actual operator coefficients. Spatially or temporally limited operators are handled by setting the applicability function to zero outside the window. The present work shows how false operator responses due to missing or uncertain data can be significantly reduced or eliminated. Perhaps the most well-known of such effects are the various 'edge effects' which invariably occur at the edges of the input data set. Examples of the performance of gradient, divergence and curl operators are given and applications to spectrum analysis is discussed. Further important applications are found in robot vision. Results showing that efficient compensation for camera saccades and elimination of visual responses due to robot arm motions can be attained. The theory is based on linear operations and is general in that it allows for both, data and operators to be scalars, vectors or tensors of higher order.< >

关键词： Filtering Multidimensional systems Convolution Robot vision systems Performance analysis Cameras Vectors Tensile stress

来源：评论

学校读者我要写书评

暂无评论

Normalized and differential convolution

Normalized and differential convolution

引用

Conference on computer vision and Pattern Recognition (CVPR)

作者： H. Knutsson C.-F. Westin Computer Vision Laboratory Department of Electrical Engineering Linköping University Linkoping Sweden

It is shown how false operator responses due to missing or uncertain data can be significantly reduced or eliminated. It is shown how operators having a higher degree of selectivity and higher tolerance against noise can be constructed using simple combinations of appropriately chosen convolutions. The theory is based on linear operations and is general in that it allows for both data and operators to be scalars, vectors or tensors of higher order. Three new methods are represented: normalized convolution, differential convolution and normalized differential convolution. All three methods are examples of the power of the signal/certainty-philosophy, i.e., the separation of both data and operator into a signal part and a certainty part. Missing data are handled simply by setting the certainty to zero. In the case of uncertain data, an estimate of the certainty must accompany the data. Localization or windowing of operators is done using an applicability function, the operator equivalent to certainty, not by changing the actual operator coefficients. Spatially or temporally limited operators are handled by setting the applicability function to zero outside the window.< >

关键词： Convolution Tensile stress Interpolation Filtering computer vision Laboratories Vectors Information analysis Signal analysis Performance analysis

来源：评论

学校读者我要写书评

暂无评论

A robust method for wideband signal separation

A robust method for wideband signal separation

引用

IEEE International Symposium on Circuits and Systems (ISCAS)

作者： V.C. Soon L. Tong Y.F. Huang R. Liu Laboratory for Image and Signal Analysis (LISA) Department of Electrical Engineering University of Notre Dame Notre Dame IN USA Department of Electrical and Computer Engineering West Virginia University Morgantown WV USA

A novel approach to solving the problem of signal separation under model uncertainties and unknown source signal characteristics is proposed. The approach features the incorporation of blind identification with clustering techniques. The approach is capable of estimating source locations and source signals under various uncertain conditions including unknown sensor gains, unknown combinations of near-field and far-field sources, unknown combinations of wideband and narrowband sources, unknown source spectral characteristics (their spectra may be overlapping or non-overlapping), and unknown number of signals.< >

关键词： Robustness Wideband Source separation Sensor arrays Sensor phenomena and characterization Uncertainty Direction of arrival estimation Narrowband Phased arrays Speech enhancement

来源：评论

学校读者我要写书评

暂无评论

Scene reconstruction and description: geometric primitive extraction from multiple viewed scattered data

Scene reconstruction and description: geometric primitive ex...

引用

Conference on computer vision and Pattern Recognition (CVPR)

作者： P. Hebert D. Laurendeau D. Poussart Computer Vision and Digital Systems Laboratory Department of Electrical Engineering Laval University QUE Canada

Robust extraction of surface parameters from multiple view scattered and noisy 3-D measurements is a delicate task. It is shown that a stable local surface description can be extracted on sections where measurement constraints are redundant with respect to a polynomial model. A segmentation approach is developed to identify these sections. The approach is based on a measurement error model which takes into account the sensor's viewpoint. An application of the approach to the extraction of straight line sections from single scan 3-D surface profiles is presented.< >

关键词： Layout Surface reconstruction image reconstruction Data mining Scattering parameters Q measurement Position measurement Signal reconstruction Sampling methods computer vision

来源：评论

学校读者我要写书评

暂无评论

Hexagonal sensor with imbedded analog image processing for pattern recognition

Hexagonal sensor with imbedded analog image processing for p...

引用

Custom Integrated Circuits Conference (CICC)

作者： M. Tremblay M. d'Anjou D. Poussart Computer Vision and Digital Systems Laboratory Department of Electrical Engineering Laval University QUE Canada

The authors present a multimodule focal plane processing sensor architecture which provides high-resolution (up to 512 /spl times/ 512 pixels) multiscale real-time analog edge extraction for robot vision. The hexagonal CMOS sensor uses a multiport addressing architecture of the pixel array to apply external multiscale analog spatial convolution followed by edge detection. The sensor architecture and its peripheral analog filtering modules are described, and results obtained from a 256 /spl times/ 256 prototype are presented. This analog satellite processing approach may be extended to various types of computational sensors, including a 3-D rangefinder, motion sensors, and tactile perception devices.

关键词： image sensors image processing Sensor arrays image edge detection computer architecture Tactile sensors Robot sensing systems Robot vision systems CMOS image sensors Convolution

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：