检索结果-内蒙古大学图书馆

A New Tracking System for Three magnetic Objectives

IEEE TRANSACTIONS ON mAGNETICS 2010年第12期46卷 4023-4029页

作者： Yang, Wan'an Hu, Chao li, mao meng, max Q. -H. Song, Shuang Chinese Acad Sci Inst Comp Technol Shenzhen 518067 Guangdong Peoples R China Chinese Acad Sci Shenzhen Inst Adv Technol Shenzhen 518067 Guangdong Peoples R China Chinese Acad Sci Shenzhen Inst Adv Integrat Technol Shenzhen 518067 Guangdong Peoples R China Chinese Acad Sci Grad Sch Shenzhen 518067 Guangdong Peoples R China Chinese Univ Hong Kong Dept Elect Engn Shatin Hong Kong Peoples R China

We have implemented a new noninvasive multiobjective tracking system, which can be used for localization of an endoscope and monitoring of heart valve prostheses and gastrointestinal transit of solid oral dosage forms or nutrients. The marker is modeled as a magnetic dipole, and the magnetic field at some point is regarded as summation of those from three dipoles. By minimizing the squared errors of magnetic field values between measurements and calculation using a hybrid of the particle swarm optimization (PSO) algorithm and the clone algorithm, an iterative result can be obtained, which is taken as the initial guess of the levenberg-marquardt (l-m) optimization method, and the first point can be determined. Subsequently, the previous computed point is used as the next initial guess of l-m algorithm, and the successive points are calculated. The tracking results demonstrate that the average position error for three objectives is 3.7 mm and the average orientation error is 1.8 degrees when the objectives move randomly in the space surrounded by the sensor array.

关键词： Dipole hybrid algorithm of PSO and clone scheme l-m algorithm magnetic marker multiobjective tracking sensor array

来源：评论

学校读者我要写书评

暂无评论

Anchovy (Engraulis ringens) and sardine (Sardinops sagax) abundance forecast off northern Chile: A multivariate ecosystemic neural network approach

引用

PROGRESS IN OCEANOGRAPHY 2010年第1-4期87卷 242-250页

作者： Yanez, Eleuterio Plaza, Francisco Carlos Gutierrez-Estrada, Juan Rodriguez, Nibaldo Barbieri, m. A. Pulido-Calvo, Inmaculada Borquez, Cinthya Pontificia Univ Catolica Valparaiso Fac Recursos Nat Escuela Ciencias Mar Valparaiso 1480 Chile Univ Huelva Palos De La Frontera 21819 Huelva Spain

An evaluation of the performance of artificial neural networks (ANNs) to forecast monthly anchovy (Engraulis ringens) and sardine (Sardinops sagax) catches in northern Chile (18 degrees 21'S-24 degrees S) is presented, using environmental variables, anchovy and sardine CPUE, fishing effort and catches between 1963 and 2007. An analysis of previous data was carried out, consisting of a non-linear cross-correlation analysis to estimate the time lags for each input variable. A multi-layer perceptron architecture model was used, calibrated with the levenberg-marquardt algorithm, thus obtaining anchovy landings and sardine CPUE forecast models. An ecosystemic approach was conducted for both models, considering local and global environmental variables, the anthropogenic effect, and the interaction between species as inputs. The variance explained by both models was slightly higher than 82% and the standard error of prediction was lower than 45%. The strong correlation between the estimated and observed series on the anchovy and sardine models suggests that ANN models capture the trend of the historical data. Furthermore, the generalization capacity along with the sensitivity analysis allowed the identification of high-weight variables in the model, as well as the partial interpretation of the statistical functional relationships between the input variables and the abundance. (C) 2010 Elsevier ltd. All rights reserved.

关键词： Artificial neural networks catches l-m algorithm Sardinops sagax Input variables ANCHOVY Engraulis ringens Sardine

来源：评论

学校读者我要写书评

暂无评论

Statistical Analysis of the Distribution of the Antibody levels to mycobacterium bovis Antigenes for Bovine Tuberculosis Diagnostics

引用

CYTOlOGY AND GENETICS 2010年第5期44卷 280-285页

作者： Redchuk, T. A. Korotkevich, N. V. Kaberniuk, A. A. Oliinyk, E. S. labyntsev, A. Yu. Romaniuk, S. I. Kolibo, D. V. Busol, V. A. Komisarenko, S. V. Natl Acad Sci Ukraine AV Palladin Biochem Inst Kiev Ukraine Natl Univ Bioresources & Nat Management Ukraine Kiev Ukraine

The antibody responses to the recombinant analogs of mycobacterium bovis antigens rmPB63 and rmPB83, as well as to tuberculin PPD, were determined in 96 bovine blood serum specimens. A statistical approach based on the approximation with multiple Gaussian curves with levenberg-marquardt algorithm was proposed for analysis of the distribution of antibody levels to the studied antigens. The results demonstrate that the recombinant mPB63 and mPB83 antigens along with the traditionally used PPD are appropriate for designing the test kits for diagnosing bovine tuberculosis at a herd level.

关键词： mycobacterium bovis approximation Statistical analysis Anti-IgA assay Data analysis Specific antibody measurement Statistical methods Purified Protein Derivative of Tuberculin bovine tuberculosis antibody response Antibodies antigens l-m algorithm Test Kit

来源：评论

学校读者我要写书评

暂无评论

Study on the causes of hypertension with improved BP neural network

Study on the causes of hypertension with improved BP neural ...

引用

2010 International Conference on E-Health Networking, Digital Ecosystems and Technologies

作者： Xiuying Dong Wang ping/Xihua University Electric and Information College Xihua University Chengdu, China

An improved neural network based on l-m (levenberg-marquard) algorithm neural network has been applied to the model for the analysis of factors on Hypertension. It can remedy the shortcoming of the slow convergence rate of traditional BP algorithm neural network. This model can determine which factors are the main reasons for high blood pressure. We have adopted the lattice fuzzy close-degree assessment and expert scoring method which quantified the various data of the factors on high blood pressure. We used the matlab to program and simulate. The results showed that: the improved BP neural network, can determine the main factors which causing the high blood pressure correctly, the error between the Predictive value and the actual value is very small. it reached the desired goal.

关键词： l-m algorithm BP neural network Close-degree grid Hypertension

来源：评论

学校读者我要写书评

暂无评论

Study on the causes of hypertension with improved BP neural network

Study on the causes of hypertension with improved BP neural ...

引用

2010 International Conference on E-Health Networking,Digital Ecosystems and Techonlogies(2010电子健康网络、数字生态系统和技术国际会议 EDT 2010)

作者： Xiuying Dong Wang ping Xihua University Electric and Information College Xihua University Chengdu China

ISBN: (纸本)9781424455140;9781424455157

An improved neural network based on l-m (levenbergmarquard) algorithm neural network has been applied to the model for the analysis of factors on Hypertension. It can remedy the shortcoming of the slow convergence rate of traditional BP algorithm neural network. This model can determine which factors are the main reasons for high blood pressure. We have adopted the lattice fuzzy close-degree assessment and expert scoring method which quantified the various data of the factors on high blood pressure. We used the matlab to program and simulate. The results showed that: the improved BP neural network, can determine the main factors which causing the high blood pressure correctly, the error between the Predictive value and the actual value is very small, it reached the desired goal.

关键词： l-m algorithm BP neural network Close-degree grid Hypertension BP neural networks Hypertension l-m algorithm mATlAB

来源：评论

学校读者我要写书评

暂无评论

Research and Development of lm Neural Network Prediction System for Grind-Hardening

Research and Development of LM Neural Network Prediction Sys...

引用

15th Conference of Abrasive Technology in China

作者： Pan, Zhongfeng Wang, Guicheng Hua, Chonglue Pei, Hongjie Jiangsu Univ Inst Precis Engn Zhenjiang 212013 Jiangsu Peoples R China

An improved neural network based on l-m algorithm has been applied to the prediction of the grind-hardening parameters against to the slow convergence rate of conventional BP neural network. And the the neural network model for grind-hardening is established. The neural network prediction system for grind-hardening process has been developed based on l-m algorithm. The functions of system is analyzed, particularly and some pivotal technology to realize the system are put forward.

关键词： Grind-hardening Neural network l-m algorithm Prediction

来源：评论

学校读者我要写书评

暂无评论

Extraction of Parameters of Random Self-Affine Fractal Surfaces from light Scattered Intensity by levenberg-marquardt algorithm

引用

Chinese Physics letters 2009年第3期26卷 122-125页

作者：张宁玉腾树云宋洪胜刘桂媛程传福 College of Physics and Electronics Shandong Normal University Jinan 250014 School of Science Shandong Jianzhu University Jinan 250101

A method for simultaneously extracting the parameters of self-affine fractal surfaces from a single experimental profile of scattered intensity data is proposed. The levenberg-marquardt algorithm is introduced to fit the theoretical equation for the scattering intensity profile to the experimental data. A precision system is designed for acquisition of scattering intensity data using the Boxcar integration technique. The surface parameters extracted （root-mean-square roughness w, lateral correlation length ζ, and roughness exponent α） are compared to those obtained using atomic force microscopy.

关键词： ROUGH SURFACES Rough surfaces Scattering light scattering coarseness index parameters Random l-m algorithm Fractals

来源：评论

学校读者我要写书评

暂无评论

Automatic computation of fundamental matrix together with interest points in correspondence between two images

Automatic computation of fundamental matrix together with in...

引用

7th International Conference on machine learning and Cybernetics

作者： Zhao, Jie Yang, Feng liu, Wei-Jing Hebei Univ Coll Elect & Informat Engn Baoding 071002 Peoples R China

ISBN: (纸本)9781424420957

The fundamental matrix is a basic tool in the analysis of scenes taken with two uncalibrated cameras. As was known, the corresponding cameras matrix may then be determined, up to a projective ambiguity, from the computed F, so the following actions will be continued such as reconstruction and so on. In the paper, the method for computing F from a set of corresponding image points is described, and the scale-invariant features extracting is used for computing the interest points. Experimental results show that, the method described is very exactly.

关键词： two-view geometry epipolar geometry fundamental matrix feature detection l-m algorithm RANSAC

来源：评论

学校读者我要写书评

暂无评论

A new method of 3d reconstruction baseing on images sequence using hand-hold camera

A new method of 3d reconstruction baseing on images sequence...

引用

7th International Conference on machine learning and Cybernetics

作者： Zhao, Jie Zhang, Ying-Kang Yan, Dong-ming Hebei Univ Coll Elect & Informat Engn Baoding 071002 Peoples R China

ISBN: (纸本)9781424420957

3D reconstruction is an important technique in the field of computer vision. The traditional 3D reconstruction technology involves the self-calibration method basing on dual absolute quadric. However, sonic drawbacks in this method cause this technology less robust. First, usually the solution of the dual quadric matrix can not satisfy the positive-definiteness condition caused by loud noise in data;second, if the motion of the camera is too restricted, the calibration degeneracy will turn up causing the reconstruction failed. To these questions, we design a new method for calibrating the camera using the camera pure rotating calibration and decomposing the essential matrix. So, the metric structure can be got directly avoiding the less robustness caused by self-calibration. Experiments show that, using this method the reconstruction has an efficient result and the procedure is convenient.

关键词： computer vision 3D reconstruction metric structure camera self-calibration camera rotating calibration essential matrix l-m algorithm bundle adjustment

来源：评论

学校读者我要写书评

暂无评论

AUTOmATIC COmPUTATION OF FUNDAmENTAl mATRIX TOGETHER WITH INTEREST POINTS IN CORRESPONDENCE BETWEEN TWO ImAGES

AUTOMATIC COMPUTATION OF FUNDAMENTAL MATRIX TOGETHER WITH IN...

引用

2008 International Conference on machine learning and Cybernetics(2008机器学习与控制论国际会议)

作者： JIE ZHAO FENG YANG WEI-JING lIU The College of Electronic and Information Engineering Hebei University Baoding 071002 China

The fundamental matrix is a basic tool in the analysis of scenes taken with two uncalibrated cameras. As was known, the corresponding cameras matrix may then be determined, up to a protective ambiguity, from the computed F, so the following actions will be continued such as reconstruction and so on. In the paper, the method for computing F from a set of corresponding image points is described, and the scale-invariant features extracting is used for computing the interest points. Experimental results show that, the method described is very exactly.

关键词： Two-View Geometry Epipolar Geometry Fundamental matriz Feature detection l-m algorithm RANSAC

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：