基于漏区边界曲率分析的射频RESURF LDMOS耐压与导通电阻优化

分析了漏区边界曲率半径与射频RESURF LDMOS击穿电压的关系,指出漏区边界的弯曲对RESURF技术的效果具有强化作用.理论分析与模拟结果表明,满足RESURF条件时,提高漂移区掺杂浓度或掺杂深度的同时相应减小漏区边界的曲率半径,可以在维持击穿电压不变的前提下,明显降低导通电阻.     

采用全对称测试结构的超薄单晶硅薄膜热导率测亮

改进了传统稳态加热法的测试结构,设计了带隔离槽的全对称悬空薄膜测试结构,并使用有限元工具对测试结构进行了优化.测量了室温下50和80nm厚度的单晶硅薄膜的横向热导率,分别为32和38W/(m·K),其相对体硅热导率(148W/(m·K))有明显下降,实验结果与BTE(Boltzmann transport equation)的理论预测曲线吻合得很好.     

场发射显示技术的发展状况

介绍了微尖阵列、碳纳米管、表面传导、弹道式电子发射、薄膜内场致发射以及类金剐石薄膜等几种场发射显示技术的发射特性，分析了场发射的工作机理。讨论了国内外场发射技术发展状况与它们在技术、材料、工艺等方面的瓶颈。展望了几种场发射显示器的市场产业化前景以及加快我国在场发射技术方面的研究步伐的建议。     

快速热处理对重掺As硅单晶中氧沉淀的影响

对重掺As硅片进行快速热处理,发现重掺As硅片中氧沉淀行为与快速热处理温度、保温时间和降温速度有很大的关系.随着快速热处理温度的升高、降温速度的增大和保温时间的延长,氧沉淀的密度增大.最后对影响的机理进行了讨论.     

Automatic image hanging protocol for chest radiographs in PACS.

Chest radiography is one of the most widely used techniques in diagnostic imaging. It comprises at least one-third of all diagnostic radiographic procedures in hospitals. However, in the picture archive and communication system, images are often stored with the projection and orientation unknown or mislabeled, which causes inefficiency for radiologists' interpretation. To address this problem, an automatic hanging protocol for chest radiographs is presented. The method targets the most effective region in a chest radiograph, and extracts a set of size-, rotation-, and translation-invariant features from it. Then, a well-trained classifier is used to recognize the projection. The orientation of the radiograph is later identified by locating the neck, heart, and abdomen positions in the radiographs. Initial experiments are performed on the radiographs collected from daily routine chest exams in hospitals and show promising results. Using the presented protocol, 98.2% of all cases could be hung correctly on projection view (without protocol, 62%), and 96.1% had correct orientation (without protocol, 75%). A workflow study on the protocol also demonstrates a significant improvement in efficiency for image display.     

A new way for multidimensional medical data management: volume of interest (VOI)-based retrieval of medical images with visual and functional features.

The advances in digital medical imaging and storage in integrated databases are resulting in growing demands for efficient image retrieval and management. Content-based image retrieval (CBIR) refers to the retrieval of images from a database, using the visual features derived from the information in the image, and has become an attractive approach to managing large medical image archives. In conventional CBIR systems for medical images, images are often segmented into regions which are used to derive two-dimensional visual features for region-based queries. Although such approach has the advantage of including only relevant regions in the formulation of a query, medical images that are inherently multidimensional can potentially benefit from the multidimensional feature extraction which could open up new opportunities in visual feature extraction and retrieval. In this study, we present a volume of interest (VOI) based content-based retrieval of four-dimensional (three spatial and one temporal) dynamic PET images. By segmenting the images into VOIs consisting of functionally similar voxels (e.g., a tumor structure), multidimensional visual and functional features were extracted and used as region-based query features. A prototype VOI-based functional image retrieval system (VOI-FIRS) has been designed to demonstrate the proposed multidimensional feature extraction and retrieval. Experimental results show that the proposed system allows for the retrieval of related images that constitute similar visual and functional VOI features, and can find potential applications in medical data management, such as to aid in education, diagnosis, and statistical analysis.     

基于三维凹包学习算法的海面小目标检测方法

对于特征类的海面小目标检测方法,现有3特征检测器通常采用凸包分类算法完成检测。在实际应用时发现,该分类算法生成的判决区域在某些情况下不能很好地反映海杂波样本集合在特征空间中的分布情况,进而对检测器性能造成一定程度的损失。相比之下,使用凹包算法生成的判决区域是由凸包内剖得到的,它能更加贴合海杂波样本的分布,因此该文将判决区域的形式由凸包转化为凹包,并在此基础之上提出一种基于3维凹包学习算法的海面小目标检测方法。同时,针对现有3维凹包算法存在的内剖效率低、无法实现恒虚警检测的缺点,该文通过优化内剖点选择方法、增加“外补”环节的方式对算法进行改进。最后,经实测CSIR数据及X波段试验雷达数据共同验证,在其他参数均相同时,该文方法的检测性能要优于已有的多特征检测方法,并且通过对凹包算法的复杂度分析证明了所提方法的应用潜力。     

矩形窗正交时频空检测算法

针对正交时频空(Orthogonal Time Frequency Space, OTFS)调制系统采用矩形窗函数时,信道矩阵结构复杂导致的鲁棒性差的问题,提出了一种基于时域处理和酉近似消息传递的检测算法。该算法首先添加循环前缀,将时域信道转换为分块对角矩阵;然后应用酉变换和近似消息传递建立迭代检测算法。仿真结果表明,所提检测算法能够在不增加复杂度的条件下有效提升检测精度和鲁棒性,特别是存在信道编码的条件下表现出2 dB的性能增益,使得该算法更适用于杂散多径、高速移动等环境,具有较高的应用价值。     

基于LSTM神经网络的压电执行器位移迟滞建模

该文搭建迟滞测量实验平台,测量一种用于LED晶圆检测压电执行器的迟滞效应,设计了一种基于长短期记忆(LSTM)神经网络的压电迟滞模型,使用时间序列预测法对压电执行器位移迟滞效应建模。将该模型与传统的Prandtl-Ishlinskii(PI)模型进行对比。实验结果表明,神经网络模型具有较好、较广泛的迟滞建模效果,对于正弦波,位移预测精度保证小于2%;对于衰减正弦波,位移预测精度可保证小于3%。较高的模型预测精度为使用压电执行器进行LED晶圆检测提供了依据。     

基于非对称监督深度离散哈希的图像检索

哈希广泛应用于图像检索任务。针对现有深度监督哈希方法的局限性,该文提出了一种新的非对称监督深度离散哈希(ASDDH)方法来保持不同类别之间的语义结构,同时生成二进制码。首先利用深度网络提取图像特征,根据图像的语义标签来揭示每对图像之间的相似性。为了增强二进制码之间的相似性,并保证多标签语义保持,该文设计了一种非对称哈希方法,并利用多标签二进制码映射,使哈希码具有多标签语义信息。此外,引入二进制码的位平衡性对每个位进行平衡,鼓励所有训练样本中的–1和+1的数目近似。在两个常用数据集上的实验结果表明,该方法在图像检索方面的性能优于其他方法。