关键词： 地铁车辆   模式识别技术   图像自动识别   检测系统  

摘要： 地铁车辆在运行过程中产生的疲劳缺陷、变形、异物掉落、部件缺失等车体异常状况的检查,现阶段主要检查方式为人工目视。这种检查方式受人员检查水平、检测手段、段场股道空间等条件限制,检查效果、效率均有待提高。基于模式识别技术的地铁车辆图像检测系统,采用隐式形状模型算法(ISM),HTM神经网络算法及支持向量机算法(SVM),实现地铁车辆各关键部件异常的图像自动识别,并研究出一套适用于地铁车辆的车体外观自动化检测系统,为车辆的自动化检查奠定基础。

关键词： 分布式光纤振动传感   多尺度一维卷积神经网络   相位敏感光时域反射   振动事件识别   模式识别  

摘要： 针对相位敏感光时域反射(Φ-OTDR)分布式光纤振动传感系统如何对振动事件进行高效准确识别的问题,本文提出了一种基于多尺度一维卷积神经网络(MS 1-D CNN)的振动事件识别方法。该方法将原始振动信号经过预加重、归一化和谱减降噪的预处理操作后得到的一维信号,直接通过MS 1-D CNN实现端到端的振动信号特征的提取和识别。MS1-DCNN在提取入侵振动信号特征时可兼顾信号时间和频率尺度,利用全连接层(FClayer)和Softmax层完成最终的识别过程,与二维卷积神经网络(2-D CNN)和一维卷积神经网络(1-D CNN)相比减少了待定参数数量。对破坏、敲击和干扰三类目标振动事件的光纤振动传感信号识别结果表明,MS 1-D CNN的识别正确率与2-D CNN相近,达到了96%以上,而处理速度提升一倍,在保持识别性能的前提下,有利于提高振动事件识别的实时性。

关键词： 模式识别   人脸识别   字典学习   局部一致性   相干性  

摘要： 字典学习算法被广泛的用于人脸识别。字典的局部一致性及相干性对于字典学习很重要,然而当前很多字典学习算法都没有考虑到字典间的局部一致性,更没有同时将二者用于字典学习,致使学习到的字典的鉴别力不是特别强,进而使得它们的识别性能不是非常理想。针对这个问题,本文提出了一种基于局部一致性和相干性的字典学习算法,并将它用于人脸识别。本算法通过构造局部一致项和相干项学习具有很强鉴别力的字典,并利用学习到的字典重构样本,最终利用重构样本与测试样本间的残差完成分类任务,在AR,ORL和FEI人脸数据库上的实验结果表明本文提出的算法取得了不错的识别效果。

关键词： Surface electromyography (sEMG)   myoelectric prosthesis   adaptive classifier   online update  

摘要： Traditional myoelectric prostheses that employ a static pattern recognition model to identify human movement intention from surface electromyography (sEMG) signals hardly adapt to the changes in the sEMG characteristics caused by interferences from daily activities, which hinders the clinical applications of such prostheses. In this paper, we focus on methods to reduce or eliminate the impacts of three types of daily interferences on myoelectric pattern recognition (MPR), i.e., outlier motion, muscle fatigue, and electrode doffing/donning. We constructed an adaptive incremental hybrid classifier (AIHC) by combining one-class support vector data description and multiclass linear discriminant analysis in conjunction with two specific update schemes. We developed an AIHC-based MPR strategy to improve the robustness of MPR against the three interferences. Extensive experiments on hand-motion recognition were conducted to demonstrate the performance of the proposed method. Experimental results show that the AIHC has significant advantages over non-adaptive classifiers under various interferences, with improvements in the classification accuracy ranging from 7.1% to 39% (p < 0.01). The additional evaluations on data deviations demonstrate that the AIHC can accommodate large-scale changes in the sEMG characteristics, revealing the potential of the AIHC-based MPR strategy in the development of clinical myoelectric prostheses.

摘要： Motor vehicle collisions cause substantial mortality, morbidity, and expense worldwide. Certain types of injuries are more likely to result from frontal versus side-impact collisions, and knowledge of these specific patterns and why they occur aids in accurate and efficient diagnosis of traumatic injuries. Although the proper use of seat belts decreases crash-related mortality during frontal impact, certain injury patterns to the torso are directly attributed to restraint use. The spectrum of seat belt-related injuries ranges from mild skin and soft-tissue contusions to traumatic bowel injuries and unstable spine injuries that require surgery. Impact with the steering wheel or windshield during a frontal crash can cause characteristic injuries to the head, neck, torso, and distal upper extremity. Steering wheel deformity is an independent predictor of serious thoracic and abdominal injury among front-seat passengers. Impact of a flexed knee with the dashboard during a frontal collision can cause knee, thigh, and hip injuries. Distal lower extremity injuries are encountered frequently when the floorboard is driven into the foot. Lateral impact crashes often result in traumatic brain, thoracic, abdominal, and pelvic injuries, which are more often fatal to occupants on the side of the impact. The specific mechanism-based injury patterns are reviewed to establish a structured systematic search pattern that enables the radiologist to identify traumatic injuries with greater accuracy and speed, thereby improving the care of patients who experience acute trauma. (C) RSNA, 2019

关键词： chILD   childhood interstitial lung disease   CT  

摘要： OBJECTIVE. T he purpose of this article is to discuss imaging techniques and a pattern-based approach for diagnosing childhood interstitial (diffuse) lung diseases in infants. CONCLUSION. Childhood interstitial (diffuse) lung disease in infants consists of a heterogeneous group of disorders previously classified with clinical, radiologic, and patho-logic features. By use of an imaging-guided algorithm, the assessment of lung volumes and the presence of ground-glass opacities or cysts can assist the radiologist in making an accurate and timely diagnosis.

关键词： Pattern recognition   binary data   sensor networks   structural health monitoring  

摘要： Recent advances in energy harvesting technologies have led to the evolution of self-powered structural health monitoring techniques that are energy-efficient. Concurrent to the emergence of self-powered sensing has been the development of power-efficient data communication protocols. One such technology is an energy-aware pulse communication architecture that employs ultrasonic pulses through the material substrate for information forwarding. This results in limited discrete binary data that raises the need for new data analysis methods for structural health monitoring purposes. A pattern recognition framework that allows interpretation of the resulting asynchronous discrete binary data for condition and damage assessment in plate-like structures is presented in this article. The proposed pattern recognition framework is based on integration of image-based pattern recognition using anomaly detection, a pattern anomaly measure, a focal density concept, and the k-nearest neighbor algorithm. Using numerical simulations, damage indication parameters were determined from the strain response of dynamically loaded plates. Simulated test cases considering different levels of damage severity, single and multiple damage regions, loading conditions, and measurement noise were studied to evaluate the effectiveness and robustness of the strategy. Furthermore, the effect of sensor density on the proposed strategy was explored. Results demonstrate satisfactory performance and robustness of the proposed pattern recognition framework for localized damage detection in plate-like structures using limited and low-resolution discrete binary data.

关键词： Pathogen recognition receptor   C-type lectin-like receptor   Toll-like receptor   Nod-like receptor   Crosstalk  

摘要： Over the last decade, invasive fungal infections have emerged as a growing threat to human health worldwide and novel treatment strategies are urgently needed. In this context, investigations into host-pathogen interactions represent an important and promising field of research. Antigen presenting cells such as macrophages and dendritic cells are strategically located at the frontline of defence against potential invaders. Importantly, these cells express germline encoded pattern recognition receptors (PRRs), which sense conserved entities from pathogens and orchestrate innate immune responses. Herein, we review the latest findings regarding the biology and functions of the different classes of PRRs involved in pathogenic fungal recognition. We also discuss recent literature on PRR collaboration/crosstalk and the mechanisms involved in inhibiting/regulating PRR signalling. Finally, we discuss how the accumulated knowledge on PRR biology, especially Dectin-1, has been used for the design of new immunotherapies against fungal infections. (C) 2018 The Authors. Published by Elsevier Ltd.

关键词： 数字图象处理   模式识别   研究  

ISBN: (纸本)9787557665708

摘要： 本书全面地介绍数字图像识别的基础理论和实用技术，以及近年来数字图像处理、图像分析与识别领域的研究成果，注重理论，突出实用。主要内容包括：数字图像处理、分析与识别的基本概念和基础理论；数字图像分割的主要方法和技术。本书结构紧凑，内容深入浅出，讲解图文并茂，可作为从事图像处理、分析和识别等相关领域的科技工作者和工程技术人员参考用书。

关键词： HVDC systems   Power system protection   DC line   Fault location   Pattern recognition   Artificial Neural Networks  

摘要： The aim of this paper is to develop a methodology based on pattern recognition of DC voltage signals for fault location in HVDC transmission systems. By using the shape similarity amongst voltage signals, an ANN is trained as a pattern recognizer to correlate post-fault DC voltage to fault distances on a DC line. The CIGRE HVDC benchmark system was used for developing and testing purposes. The HVDC system is modeled and simulated in PSCAD, while the ANN is trained and implemented in MATLAB (TM). The proposed methodology was developed and tested considering different operational conditions of the adopted HVDC system. After an exhaustive evaluation process considering different fault locations and fault resistances, the method presented here showed a very accurate and robust behavior.