Human thermal models for evaluating infrared images

Discusses comparing infrared images under various thermal environmental conditions through normalization of skin surface temperature. To evaluate IR images obtained under various thermal environmental conditions, we proposed a human thermal model with which IR images obtained under certain thermal environmental conditions can be converted into images under other conditions. The model was based on a numerical calculation of the bio-heat transfer equations that express heat transfer phenomena within the human body. A 16-cylinder-segment model was used as the geometry of the human body. Comparisons of IR images with their converted images indicate that this method is effective in eliminating the influence of the thermal environmental conditions. However, the difference between the converted images and the original ones varies among segments. In future work, we will use this method to investigate the IR images of several subjects under various thermal environments.     

Increasing the effective resolution of thermal infrared images

Thermal infrared (TIR) imaging is recognized as the most efficient technique for the study of skin temperature distribution. In specific diseases, characteristic changes can be measured from target anatomical sites. In this way, objective noninvasive investigations can be of diagnostic value. TIR imaging of the breast for breast cancer risk assessment is an example. Although TIR imaging possesses the advantages of being noninvasive, risk free, and considerably less expensive, it suffers the disadvantage of a lack of resolution due to blur compounded by rather high levels of noise. A maximum a posteriori probability (MAP) image restoration philosophy is proposed to solve this problem of resolution. The objective is three-fold: to increase the resolution of the measured image by using a type of 2:1 zooming; to remove the noise; and simultaneously to preserve the detail of features, including, in particular, the sharpness of edges     

基于平行图像的糖尿病视网膜病变智能诊断

针对深度学习诊断糖尿病视网膜病变（diabetic retinopathy, DR）面临数据集小、类别不均衡及诊断效果不佳等问题,本文提出基于平行图像和Swin Transformer的DR分级模型。首先基于StyleGAN2-ada构建平行图像生成模型,解决训练图像过少和类别失衡问题。经FID、KID和目视评估,构建的平行图像符合后续工作要求。然后,基于注意力与窗口滑动机制构建DR诊断模型改善诊断效果。最后,使用平行图像训练诊断模型。经验证,本文提出的诊断模型准确率为93.5%、特异性最高为99%、F1分数最高为0.96。与原始图像相比,使用平行图像训练模型后其准确率提升20%、精确率最高提升70%。与其他三种深度学习模型对比,本文所提方法各项指标均达到最优。以上结果表明,本文构建的模型可在小样本数据集下实现较好的诊断效果。     

单轴并联式混合动力汽车动力总成建模与仿真

以优化整车动力性和经济性为目标,在Insight混合动力汽车底盘基础上,设计以镍氢电池组和永磁同步电动机驱动的汽车动力系统.利用电辅助控制策略,在满足整车动力性能指标的基础上,使用汽车专用仿真软件ADVISOR(advanced vehicle simulator)进行二次开发,采用前向仿真与后向仿真相结合的方法,根据实际情况构建系统模型,对发动机参数、电机参数选择以及能源匹配进行了仿真研究.仿真结果表明:该单轴并联式混合动力样车的动力系统设计方案是可行的,研究结果可为混合动力电动汽车动力总成的优化设计提供理论依据.     

基于面部红外热图的恐惧情绪识别

恐惧情绪是人体应对外界刺激的一种反应,其产生会引起人体面部皮肤温度的变化。根据红外热图反映物体表面温度分布的原理,提出了一种基于面部红外热图对恐惧情绪进行识别的方法。首先采用指数函数拟合方法对现有的传热模型进行简化,将红外热像仪所采集的面部红外热图转化为血液灌注伪彩色图,寻找感兴趣区域(前额区域);然后,提取感兴趣区域的血液灌注变化曲线的相关特征(斜率、置信度、均值、标准偏差),并采用斯皮尔曼相关系数分析其与受试对象恐惧程度自评分之间的相关性;最后采用与恐惧程度自评分相关性最高的标准偏差对受试对象进行恐惧情绪的识别。实验结果表明,当受试对象产生恐惧情绪时,前额区域的血液灌注呈现出明显的下降趋势,这与现有的恐惧情绪研究结论一致;血液灌注值的标准偏差可以作为恐惧情绪识别的主要特征(阈值0.14)。采用本文所提出的方法,对28例样本进行测试,受试对象恐惧情绪的识别准确率达到85.7%,具有较高的可靠性。     

Laboratory-based coursework in parallel processing using thetransputer

The design of a low cost, target transputer system is described, along with code development facilities, that allows the concepts of parallel processing to be taught in a practical manner. Outlines of eight graded exercises are also included as an indication of how the system might be used in an undergraduate laboratory     

平面变压器中并联绕组的设计

在低压大电流DC-DC模块电源中，经常使用并联绕组。由于电磁场的高频效应 (集肤和邻近效应)，可能导致电流在各并联绕组层中不均分，产生较高的交流电阻。本文提出使电流在各并联层中均分的绕组布置方法。并通过有限元分析软件和实验证实了所提出的绕组布置方法的正确性和有效性。     

Application of parallel genetic algorithms to generation expansion planning using parallel processors

This paper presents an application of parallel genetic algorithms (PGA) to the optimal long-range generation expansion planning. The problem can be formulated as a combinatorial optimization problem that determines the order of introduced generation units at each interval of the year. The proposed method considers introduced power limits of each technology, maximum loads at each interval, and load duration curves at each interval. Appropriate string representation for the problem is presented. Binary and decimal coding and three selection methods are compared. The method is developed on a transputer that is one of the parallel processors. The feasibility of the proposed method is demonstrated using a typical expansion problem with four technologies and five intervals. The method is then compared with conventional dynamic programming and a simple genetic algorithm with promising results.     

An experimental investigation of the feasibility of using silicone and gallium arsenide solar batteries on space vehicles for receiving energy of laser infrared emission

The feasibility of transmitting electric power in space to solar batteries of space vehicles via the channel of laser infrared emission is shown. Evaluation of the efficiency of solar batteries for the given type of power transmission has been made. Possible methods of optimizing the design of space solar batteries in regard to conditions of detection of energy of laser infrared emission have been discussed.     

High-order volterra series analysis using parallel computing

This paper addresses the determination of high-order Volterra transfer functions of non-linear multiport networks containing multidimensional non-linear elements. A new method is developed for utilizing parallel computing which is very efficient for approximately 10–20 processors. The utilization of each processor may be as high as 80%–95%. The developed program is very flexible as it is ANSI C++ compatible and may run on both single- and multiprocessor computers. Using about eight processors it is possible to analyse rather complicated non-linear circuits up to ninth order in a few hours. © 1997 by John Wiley & Sons, Ltd.     

Dynamic interconnection component using wireless infrared technology

An efficient dynamic interconnection model using wireless infrared technology and the theory of optical interconnections was constructed to design a dual-channel interconnection component. There were three conditions between the rotating optical field and the stationary optical field: end separation, angle misalignment and lateral misalignment. The calculation formulas were given for these three conditions. A dual-channel optical interconnection component was designed based on the dynamic interconnection model and the data transmission rate of the component was measured. The experimental result showed that the dual-channel optical interconnection component could transmit optical signals across the rotating interface. The maximum transmission rate can reach 2.14 Mb/s. Translated from Journal of Tianjin University, 2006, 39(8): 985–988 [译自: 天津大学学报]     

一种采用电池组并联供电的直流系统

提出了由多组电源模块配套12 V蓄电池并联供电的直流系统方案,与传统电池串联供电的直流系统有着明显的优点,通过软硬件阐述了该方案中电源模块对蓄电池的智能充放电管理,实验结果表明,此方案下的系统令蓄电池的寿命更长,维护成本更低,系统更加稳定。     

Protection of parallel transmission lines using wavelet transform

A new scheme to enhance the solution of the problems associated with parallel transmission line protection is presented in this paper. The scheme depends on the three line voltages and the six line currents of the two parallel lines at each end. Fault detection, fault discrimination, and calculation of the phasors of the measured signals are done by using wavelet transform (WT). By comparing the magnitudes of the estimated current phasors of the corresponding phases on both lines, internal faults on the parallel lines can be identified. Also, by calculating the distance element of the phases on which a disturbance is detected and having a very small current difference magnitude can enhance and strengthen the scheme. Studies show that all types of faults at different loading conditions can be correctly identified in less than one cycle of the fundamental frequency.     

Optoelectronic parallel computing using optically interconnectedpipelined processing arrays

We describe our research on optoelectronic parallel computing systems. Our architecture is based on a multilayer pipeline of two-dimensional optoelectronic device arrays in which each pixel is composed of an optical input channel, a general purpose programmable processor, local memory, and a surface-emitting laser diode as an optical output channel. Free-space optics provides parallel, global communication between layers in the pipeline via optical paths which are dynamically reconfigurable. Demonstration systems and some applications are described     

Recognizing the ID of modulated LED tube lights by using camera motion blur

In this paper, we present an image‐based ID recognition method for ID‐modulated light‐emitting diode (LED) tube lights by using the motion blur captured by a moving camera. This method is applied to a novel camera‐based indoor positioning system, which can provide exact location for mobile users. In this system, high‐intensity LED tubes are used concurrently as the illumination devices and optical markers. The flashing of each LED lamp is modulated, and the entire tube expresses an ID message, which can be captured by a normal camera installed in a mobile terminal. The flashing occurs at a high frequency and without degrading the illumination function. However, when the exposure time of the camera is longer than the flicker period of the LED lamps, it is difficult to capture the ID pattern. We propose a method that uses motion blur to overcome this limitation. During the period of exposure, if the user manually shakes the camera in the proper direction, a streaked pattern is developed on the captured image frame, which can be used for retrieving an ID number. Moreover, we can also obtain position estimation of the terminal from motion‐blurred images. Experimental results show that with careful operation it is feasible to recognize the ID of LED tubes successfully. © 2012 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.     

Removal of adherent noise in images by using multiple cameras

In this paper, we propose a new method for removing view‐disturbing noise from images taken with multiple cameras. In outdoor environments, scenes taken by cameras are often hard to see because of adherent noise on the surface of the lens‐protecting glass. The proposed method analyzes multiple camera images describing the same scene, and synthesizes an image in which adherent noise is eliminated. We show the effectiveness of the proposed method through experimental results. © 2008 Wiley Periodicals, Inc. Electr Eng Jpn, 164(3): 50–59, 2008; Published online in Wiley InterScience ( www.interscience.wiley.com ). DOI 10.1002/eej.20697     

利用电动汽车可调度容量辅助电网调频研究

电动汽车既可作为可控性负荷,也可作为分布式电源,能够为电网提供可调度容量,参与调频等辅助服务。但该可调度容量受用户出行需求及电池损耗等因素的制约,不能无限制地调度。基于此,对电动汽车采取"分散接入,集中控制"的管理模式,首先基于用户出行需求及电池使用寿命等约束,对电动汽车可调度容量进行评估。进而建立了计及可调度容量的电动汽车集中管理器充放电静态频率特性模型,以单区域系统为例模拟了电动汽车参与负荷调频的作用效果。仿真结果表明,利用电动汽车可调度容量辅助电网调频,不仅可以快速有效地减小系统频率偏差,提高电能质量,还能减小传统调频机组的备用容量,进而提高电网经济性。研究电动汽车参与调频的作用效果时,用户需求及电池损耗是不容忽视的影响。     

Realization of 3-degree-of-freedom control using fine-grain parallel processing

The article reports a sprout of 3-degree-of-freedom (DOF) control. The limitation of feedback control, including 2-DOF control, has been raised for a long time; however, a suitable control theory for overcoming those limitations has still not been formulated. An underlying reason is that the disturbance-suppression performance and noise sensitivity cannot be decoupled. Several researches have tried to tackle this problem using a hardware approach, as a noise level depends on a hardware configuration. It should be noted that the hardware design expands spatiotemporal resolution of a system. This approach helps in reducing noise, namely, it works as the 3rd-DOF for a control system. Therefore, an improvement in the hardware design could be a new angle for solving a mixed-sensitivity problem. This article quantitatively presents the relation between noise-reduction performance and the spatiotemporal resolution, and provides a foothold for the 3-DOF control.     

High-efficiency red and infrared light-emitting diodes using radialoutcoupling taper

In this paper, we give an overview of light-emitting diodes (LEDs) with radial tapers. Light is generated in the very center of a circularly symmetrical structure and is outcoupled at a tapered ring. Encapsulated devices with an emission wavelength of 980 nm achieve wallplug efficiencies of 48%. Non-encapsulated InGaAlP-based red-emitting LEDs show quantum efficiencies of 13%. A new device design combines the taper with a wafer-scale soldering technique promising a feasible fabrication method     

基于灰度冗余和SURF算法的电气设备红外和可见光图像配准

为了对电力设备故障进行更为有效和直观的诊断,提出了一种基于灰度冗余和SURF算法的红外和可见光图像的配准方法.该算法对红外图像和可见光图像进行基于灰度冗余的处理,利用SURF方法分别找到两幅图像的特征点并求出其描述子,通过RANSAC(随机采样一致性)对这些特征点进行精确配对,通过最小二乘法求得仿射变换参数并实现两幅图像的配准.实验表明,该方法与其他配准方法相比,具有速度快,鲁棒性高等优点,具有较高的实用价值.