检索结果-南通市图书馆

A distribution prior model for airplane segmentation without exact template

在线全文

学校读者我要写书评

暂无评论

Journal of Systems Engineering and Electronics 2020年第1期31卷 56-63页

作者： DAI Ming zhou zhiheng GUO Yongfan School of Electronics and Information Engineering South China University of Technology

In many practical applications of image segmentation problems,employing prior information can greatly improve segmentation *** paper continues to study one kind of prior information,called prior *** this research,there is no exact template of the object;instead only several samples are *** proposed method,called the parametric distribution prior model,extends our previous model by adding the training procedure to learn the prior distribution of the *** this paper establishes the energy function of the active contour model(ACM)with consideration of this parametric form of prior ***,during the process of segmenting,the template can update itself while the contour *** are performed on the airplane data *** results demonstrate the potential of the proposed method that with the information of prior distribution,the segmentation effect and speed can be both improved efficaciously.

关键词： image segmentation active contour model(ACM) prior distribution level set method

Construction of compact RBF network by refining coarse clusters and widths

在线全文

学校读者我要写书评

暂无评论

Journal of Systems Engineering and Electronics 2009年第6期20卷 1309-1315页

作者： Zeng Delu zhou zhiheng Xie Shengli School of Electronic and Information Engineering South China Univ. of Technology Guangzhou 510641 P. R. China

It is known that centers, widths, and weights are three mainly considered factors in constructing a radial basis function（RBF） *** paper aims at constructing a compact RBF network with two main *** the first step, the coarse clusters computed from triangle inequalities are refined to obtain the locations of centers by the defined maximum degree spanning tree（MDST）.Meanwhile the coarse widths are *** the second step, a learning algorithm referred to as anisotropic gradient descent method is presented to further refine the above coarse *** of the proposed algorithm show its great performance in times series prediction and classification.

关键词： clustering anisotropic gradient descent radial basis function time series prediction boundary extraction.

Global minimization of adaptive local image fitting energy for image segmentation

在线全文

学校读者我要写书评

暂无评论

Journal of Systems Engineering and Electronics 2014年第2期25卷 307-313页

作者： Guoqi Liu zhiheng zhou Shengli Xie School of Electronic and Information Engineering South China University of Technology School of Automation Guangdong University of Technology

The active contour model based on local image fitting （LIF） energy is an effective method to deal with intensity inhomo- geneities, but it always conflicts with the local minimum problem because LIF has a nonconvex energy function form. At the same time, the parameters of LIF are hard to be chosen for better per- formance. A global minimization of the adaptive LIF energy model is proposed. The regularized length term which constrains the zero level set is introduced to improve the accuracy of the bound- aries, and a global minimization of the active contour model is presented, in addition, based on the statistical information of the intensity histogram, the standard deviation σ with respect to the truncated Gaussian window is automatically computed according to images. Consequently, the proposed method improves the performance and adaptivity to deal with the intensity inhomo- geneities. Experimental results for synthetic and real images show desirable performance and efficiency of the proposed method.

关键词： image segmentation level set adaptive local imagefitting (LIF) energy.

RBF neural network and active circles based algorithm for contours extraction

在线全文

学校读者我要写书评

暂无评论

Progress in Natural Science:Materials International 2007年第6期17卷 681-686页

作者： zhou zhiheng, Zeng Delu and Xie Shengli(College of Electronic & Information Engineering, South China University of Technology, Guangzhou 510640, China) College of Electronic & Information Engineering South China University of Technology Guangzhou 510640 China

For the contours extraction from the images, active contour model and self-organizing map based approach are popular nowadays. But they are still confronted with the problems that the optimization of energy function will trap in local minimums and the contour evolutions greatly depend on the initial contour selection. Addressing to these problems, a contours extraction algorithm based on RBF neural network is proposed here. A series of circles with adaptive radius and center is firstly used to search image feature points that are scattered enough. After the feature points are clustered, a group of radial basis functions are constructed. Using the pixels’ intensities and gradients as the input vector, the final object contour can be obtained by the predicting ability of the neural network. The RBF neural network based algorithm is tested on three kinds of images, such as changing topology, complicated background, and blurring or noisy boundary. Simulation results show that the proposed algorithm performs contours extraction greatly.

关键词： contours extraction RBF neural network dynamic clustering.

Vector bundle constraint for particle swarm optimization and its application to active contour modeling

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Progress in Natural Science:Materials International 2007年第10期17卷 1220-1225页

作者： Zeng Delu, zhou zhiheng and Xie Shengli(College of Electronic and Information Engineering, South China University of Technology, Guangzhou 510641, China) College of Electronic and Information Engineering South China University of Technology Guangzhou 510641 China

Active contour modeling (ACM) has been shown to be a powerful method in object boundary extraction. In this paper, a new ACM based on vector bundle constraint for particle swarm optimization (VBCPSO-ACM) is proposed. Different from the traditional particle swarm optimization (PSO), in the process of velocity update, a vector bundle is predefined for each particle and velocity update of the particle is restricted to its bundle. Applying this idea to ACM, control points on the contour are treated as particles in PSO and the evolution of the contour is driven by the particles. Meanwhile, global searching is shifted to local searching in ACM by decreasing the number of neighbors and inertia. In addition, the addition and deletion of particles on the active contour make this new model possible for representing the real boundaries more precisely. The proposed VBCPSO-ACM can avoid self-intersection during contour evolving and also extract inhomogeneous boundaries. The simulation results proved its great performance in performing contour extraction.

关键词： boundary extraction active contour modeling particle swarm optimization velocity update vector bundle.

Regular Quasi Cyclic Low Density Parity Check Codes with Girth 8 from Elementary Number Theory

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

China Communications 2012年第4期9卷 80-88页

作者： He Guofeng Li Xiangxue Li Qiang zhou zhiheng Zheng Dong School of Electronic Information and Electrical Engineering Shanghai Jiao Tong University Shanghai 200240 P. R. China Hangzhou Key Lab of E-business and Information Security Hangzhou Normal University Hangzhou 310036 P. R. China Department of Computer Science and Technology East China Normal University Shanghai 200241 P. R. China

This paper is concerned with (3,n) and (4,n) regular quasi-cyclic Low Density Parity Check (LDPC) code constructions from elementary number *** the column weight,we determine the shift values of the circulant permutation matrices via arithmetic *** proposed constructions of quasi-cyclic LDPC codes achieve the following main advantages simultaneously:1) our methods are constructive in the sense that we avoid any searching process;2) our methods ensure no four or six cycles in the bipartite graphs corresponding to the LDPC codes;3) our methods are direct constructions of quasi-cyclic LDPC codes which do not use any other quasi-cyclic LDPC codes of small length like component codes or any other algorithms/cyclic codes like building block;4)the computations of the parameters involved are based on elementary number theory,thus very simple and *** results show that the constructed regular codes of high rates perform almost 1.25 dB above Shannon limit and have no error floor down to the bit-error rate of 10-6.

关键词： quasi-cyclic LDPC code error floor Shannon limit number theory

Self-Supervised Learning with Min-Max Distance Approach for Few-Shot Steel Surface Defects Recognition

在线全文

学校读者我要写书评

暂无评论

Self-Supervised Learning with Min-Max Distance Approach for ...

第43届中国控制会议

作者： Tianlei Wang Xinyao Chen Zeliang Li Zhencai zhou zhiheng zhou School of Electronic and Information Engineering Wuyi University School of Mechanical and Automation Engineering Wuyi University Technical DepartmentGuangdong Province Kejie Machinery Automation Company

Surface defects in steel are crucial for quality inspection in the manufacturing of industrial metal *** deep learning methods have been extensively studied in this field, the lack of large-scale annotated defect samples imposes significant labeling pressure, making training fully supervised deep models challenging. In response to this issue, we employ self-supervised learning for few-shot recognition,effectively learning representations from limited unlabeled data and achieving defect recognition with a small amount of labeled samples. In particular, we construct Other encoder with multiview information and Base encoder with single view information based on the distillation self-supervised learning framework through the multi-crops mechanism, and the Other encoders and Base encoders perform information share. Our main contribution is that, based on this framework, we propose a new regularization method to solve the problem of instances deviating from the center of self-supervised clustering by minimizing the maximum instance distance(MMD). In testing,fine-tuning the MMD pre-trained model with a small number of labeled data achieves superior recognition performance on the steel surface defect dataset NEU, significantly outperforming other testing methods.

关键词： deep learning self-supervised learning L1-norm

来源： cnki会议评论

在线全文

cnki会议

学校读者我要写书评

暂无评论

神经网络方法应用于轮廓定位

计算机工程与应用 2011年第19期47卷 179-181页

作者：肖平周智恒华南理工大学电子与信息学院广州510641

关于从图像中定位物体轮廓的问题,目前所采用的活动轮廓模型和基于自组织神经网络的算法,存在能量泛函优化容易陷入局部极值和演化过程依赖于初始轮廓的选取等问题。提出了一种基于RBF神经网络的轮廓定位算法。首先,通过自适应梯度阈值... 详细信息

关于从图像中定位物体轮廓的问题,目前所采用的活动轮廓模型和基于自组织神经网络的算法,存在能量泛函优化容易陷入局部极值和演化过程依赖于初始轮廓的选取等问题。提出了一种基于RBF神经网络的轮廓定位算法。首先,通过自适应梯度阈值方法来获取图像特征点。然后,通过特征点的聚类建立一组基函数,把图像像素点的像素值和梯度构造输入向量空间,在网络权值训练完成后,利用网络的预测功能来准确判断物体轮廓。与传统算法相比,仿真结果表明提出的轮廓定位算法可以高效地实现目标轮廓定位。

关键词：轮廓定位神经网络自适应阈值

基于扩展仿射模型的不确定性静态电压稳定性全局灵敏度分析

在线全文

学校读者我要写书评

暂无评论

电工技术学报 2021年第13期36卷 2821-2831页

作者：乐健廖小兵李奔周子恒彭学林武汉大学电气与自动化学院武汉430072 国网鄂州供电公司鄂州436000 国网武汉供电公司武汉430000

波动性可再生能源大规模并网成为电力系统新的重要不确定性因素,准确评估这种不确定性对电力系统静态电压稳定性的影响,有利于提高系统运行安全水平。计及可再生能源和负荷功率的不确定性,该文根据扩展仿射算术建立了基于L指标的静态电... 详细信息

波动性可再生能源大规模并网成为电力系统新的重要不确定性因素,准确评估这种不确定性对电力系统静态电压稳定性的影响,有利于提高系统运行安全水平。计及可再生能源和负荷功率的不确定性,该文根据扩展仿射算术建立了基于L指标的静态电压稳定性区间评估模型,利用解析化方差分解的全局灵敏度指标辨识输入变量对静态电压稳定性影响的重要程度。多个IEEE算例的仿真结果表明,该文所提方法相比常用仿射算术能更有效地抑制区间扩张效应,同时基于解析化方差分解的全局灵敏度分析方法能快速、高效地辨识出影响静态电压稳定性的关键因素。

关键词：区间不确定性全局灵敏度分析静态电压稳定性扩展仿射算术方差分解