一种基于最大熵模型的实时垃圾邮件判定及阻断技术研究

垃圾邮件过滤是网络安全领域的重大研究课题和难题.本文针对传统技术的不足,使用机器学习领域经典的最大熵模型对垃圾邮件进行判定,并结合其采用针对SMTP邮件通信协议的实时阻断技术.实验结果表明,该技术简单有效,并适合实时在线处理.     

数字全息技术在温度场测量中的应用

温度的微小变化必然会引起空气折射率的微小变化,进而影响通过该区域光的相位。采用迈克耳孙干涉仪与数字全息技术相结合的技术方案,实现对温度场分布的测量。借助数字视频技术,该技术架构可直接推广用于测量透明介质的流场分布及变化过程,如气流或水流中的密度分布以及变化过程等。     

一种分布式无人值守行发射机设计

介绍一种分布式无人值守行发射机的设计和应用.该发射机通过采用模块化、幅度均衡、温度控制等设计手段,使之具有较高的可靠性及环境适应能力.该发射机通过试用,验证了其环境适应能力和高的可靠性.     

基于Camera Link接口的图像跟踪系统的设计与实现

为了提高图像跟踪对小目标的跟踪精度,采用高帧频和高分辨率的数字相机作为输入。系统采用以DSP和FPGA以及大容量缓存的硬件架构,通过FPGA的逻辑编程灵活调配系统的资源,将图像采集、图像预处理、图像跟踪、图像显示等功能合理分配至各个硬件单元。经实验验证,各个功能单元达到预期效果。这里介绍的基于Carema Link接口的图像跟踪系统,其显著特点在于图像处理能力强,跟踪精度高。     

融合注意力机制与密集多尺度特征的异质噪声虹膜分割方法

针对当前分割方法不能兼顾可见光与近红外光谱的异质噪声虹膜图像的分割精度与效率问题,提出一种基于注意力机制与密集多尺度特征融合的编-解码网络.首先,引入基于深度可分离卷积的改进残差瓶颈单元,降低参数量与计算量的同时防止信息丢失与梯度混淆;其次,改进密集空洞空间金字塔模块的空洞率组合并放置于编码器后,以增强多尺度特征融合;...     

A 1-V, 82-dB, 2.5-MS/s,single loop,single bit delta-sigma modulator in 0.13-μm CMOS technology

It is challenging to design high speed Delta-Sigma modulator using sub-micron process with low supply voltage. Compared with multi-stage or multi-bit design, the single loop, single bit Delta-Sigma modulator has relaxed requirement for the building blocks under low-voltage operation, which make it possible to get high conversion rate by increasing sampling frequency. In this study, a low voltage, high speed 4th-order Delta-Sigma modulator using input feed-forward is presented. Implemented with 0.13-μm CMOS technology and 1.0-V supply voltage, the discrete-time Delta-Sigma modulator achieves 2.5-MS/s conversion rate and 82-dB dynamic range, with the sampling frequency of 160-MHz and OSR of 64.     

Efficient Infrared Solar Cells Employing Quantum Dot Solids with Strong Inter-Dot Coupling and Efficient Passivation

Lead chalcogenide quantum dot (QD) infrared (IR) solar cells are promising devices for breaking through the theoretical efficiency limit of single-junction solar cells by harvesting the low-energy IR photons that cannot be utilized by common devices. However, the device performance of QD IR photovoltaic is limited by the restrictive relation between open-circuit voltages (V_OC) and short circuit current densities (J_SC), caused by the contradiction between surface passivation and electronic coupling of QD solids. Here, a strategy is developed to decouple this restriction via epitaxially coating a thin PbS shell over the PbSe QDs (PbSe/PbS QDs) combined with in situ halide passivation. The strong electronic coupling from the PbSe core gives rise to significant carrier delocalization, which guarantees effective carrier transport. Benefited from the protection of PbS shell and in situ halide passivation, excellent trap-state control of QDs is eventually achieved after the ligand exchange. By a fine control of the PbS shell thickness, outstanding IR J_SC of 6.38 mA cm⁻² and IR V_OC of 0.347 V are simultaneously achieved under the 1100 nm-filtered solar illumination, providing a new route to unfreeze the trade-off between V_OC and J_SC limited by the photoactive layer with a given bandgap.     

Recent Progress of Bismuth Effect on All-Inorganic Lead-Free Metal Halide Derivatives: Crystals Structure,Luminescence Properties,and Applications

Bismuth (Bi³⁺)-included lead-free metal halide (LFMH) materials attract much attention in lighting, display, photodetectors, X-ray detectors, and photovoltaic fields, due to the tunable luminescence and optoelectronic performance in response to crystal and electronic structure, morphology, and particle sizes. This review summarizes Bi³⁺-included LFMH materials about their preparation approach, crystal and electronic structure properties, luminescence performance, and emerging applications. Notably, Bi³⁺ ions not only can act as framework cation to construct stable LFMH structure, but can also incorporate into LFMH materials as activators or sensitizers to generate remarkable luminescence tuning and band engineering. The Bi³⁺ effect on the luminescence and optoelectronic properties of LFMH materials, including, promotion of exciton localization, enhancement of light absorption in near-ultraviolet region, action as sensitizer ions to transfer energy to rare earth or transition metal ions and emission of highly-efficient light is systematically summarized. The proposed structure-luminescence relationship offers guidance for the optimization of current Bi³⁺-included LFMH materials and the exploitation of new LFMH derivatives.     

Analysis of error in the fixed-point implementation of two-dimensional discrete wavelet transforms

Computational error due to the fixed-point implementation of two-dimensional (2-D) discrete wavelet transform (DWT) is analyzed. This analysis is based on the exact knowledge of the DWT analysis and synthesis filters and the word length of the original image. In the fixed-point implementation, it is crucial to understand and analyze effects of finite precision in filters coefficients as well as rounding of intermediate calculations for the purpose of storage and/or transmission. Analyses and formulations are presented for both convolution and lifting approaches and they are validated by Monte Carlo simulations. The specific example used throughout this work is the lossy wavelet transformation used in the JPEG2000 compression standard.     

Armlets and balanced multiwavelets: flipping filter construction

In the scalar-valued setting, it is well-known that the two-scale sequences {q/sub k/} of Daubechies orthogonal wavelets can be given explicitly by the two-scale sequences {p/sub k/} of their corresponding orthogonal scaling functions, such as q/sub k/=(-1)/sup k/p/sub 1-k/. However, due to the noncommutativity of matrix multiplication, there is little such development in the multiwavelet literature to express the two-scale matrix sequence {Q/sub k/} of an orthogonal multiwavelet in terms of the two-scale matrix sequence {P/sub k/} of its corresponding scaling function vector. This paper, in part, is devoted to this study for the setting of orthogonal multiwavelets of dimension r=2. In particular, the two lowpass filters are flipping filters, whereas the two highpass filters are linear phase. These results will be applied to constructing both a family of the most recently introduced notion of armlet of order n and a family of n-balanced orthogonal multiwavelets.