Terahertz time-domain spectroscopy for explosive imaging

Terahertz (THz) imaging which is a new technology for material classification and nondestructive detection has been extensively investigated in the past decade. The time-domain waveform acquired at each point of the object by using the THz time-domain spectroscopy contains much information about the object. Processing this waveform will present the characters of the object. Several methods are adopted to generate the image of the explosive samples and results are compared and discussed. Experiment results indicate that this new imaging technology can be used for explosive detection.     

A study on the digital moiré technique with circular and radial gratings

A digital moiré technique with circular and radial gratings is proposed to measure in-plane displacement and strain distributions of soft materials in a polar coordinate system. By introducing the coordinate-transform technique, the concentric circles and radial lines in the Cartesian coordinate system are converted to cross gratings in a polar coordinate system. Then, the digital moiré fringes are generated using logical and filter operations on the one-bit binary images. Finally, the in-plane displacement distributions are obtained by means of the four step phase-shifting procedure. Also, the measurement principles and the basic procedures of the new digital moiré technique are explained in detail. An application for rubber sheet with a single-edge crack is studied, and the experimental results show that the presented method is feasible.     

图像融合效果评价方法研究

以IKONOS影像为数据源，对IHS、PC，Gram—Schmidt和基于亮度调节的平滑滤波（SFIM）四种融合算法进行融合处理，对其结果进行了评价．结果表明，定性比较主观性较强，定量比较方法虽多，但大都有一定的局限性，不同谱段的计算结果总有些差异．因此，定量评价时应该多选一些波段与评价方法才能获得比较客观的评价结果．光谱曲线的比较不失为一种比较好的方法，在对光谱曲线进行比较时本文引入了光谱角度和几何距离两种方法进行定量的比较．     

基于分形脊波神经网络的遥感图像分类研究

在分形理论和脊波神经网络的基础上，综合利用彩色遥感图像的光谱、纹理和形状特征，提出了一种彩色遥感图像的分类新方法．该方法把彩色图像的蓝、绿、红波段作为3个光谱特征，由分形理论计算的DBC维和多重分形维数作为2个纹理特征，平均不变矩作为1个形状特征，并利用对曲线具有极强方向识别能力的脊波神经网络作为分类器．实验结果表明，提出的彩色遥感图像分类方法具有较高的分类准确率和较强的抗噪音能力．     

Application of audience-seats characteristics to the sound field analysis for large enclosures

The analysis of the sound field of a hall including the effect of audience seats has difficulties in computing time and memory. Although extensive investigations have been conducted for analyzing sound fields, there is scarcely any report dealing with the sound field including this effect. The main cause may be the complex shape and amount of materials of the audience seats. However, there is a possibility that periodical grooved structures like seat rows can be replaced by an appropriate imaginary boundary with an impedance of quasi-local reaction at the top of seats. The purpose of this research is to reduce the calculation load of acoustical analysis by treating the seat rows as an imaginary-impedance boundary which is almost equivalent to the characteristics of audience seats. The SDE filter, which can express excess attenuation caused by seat rows, is proposed and some applications to both the geometrical and wave theoretical methods are presented. To examine the validity of this method, some numerical examples calculated using an image method are discussed in comparison with the measured data. Furthermore, the numerical results of its application to the boundary element method are presented and discussed in terms of its effectiveness and validity.     

Image feature extraction by dynamic neural filtering and phase-only joint transform correlation

Neural network-based image processing algorithms present numerous advantages due to their supervised adjustable properties. Among various neural network architectures, dynamic neural networks, Hopfield and Cellular networks, have been found inherently suitable for filtering applications. Combining supervised and filtering features of dynamic neural networks, this paper presents dynamic neural filtering technique based on Hopfield neural network architecture. The filtering technique has also been implemented by using phase-only joint transform correlation (POJTC) for optical image processing applications. Filtering structure is basically similar to the Hopfield neural network structure except for the adjustable filter mask and 2D convolution operation instead of weight matrix operations. The dynamic neural filtering architecture has learnable properties by back-propagation learning algorithm. POJTC presents significant advantages to achieve the operation of summing the cross-correlation of bipolar data by phase-encoding bipolar data in parallel. The image feature extraction performance of the proposed optical system is reported for various image processing applications using a simulation program.     

Frequency stabilization of multiple semiconductor lasers using digital feedback loop

We propose and demonstrate a simple technique to stabilize the frequencies of multiple semiconductor lasers simultaneously using a Fabry-Perot etalon filter and a digital signal processing. Unlike other stabilization techniques, the proposed technique could stabilize DFB lasers using only one digital feedback control loop. The result shows that the proposed technique could maintain the frequency stability of 16 DFB lasers within ±150 MHz.     

Measurement of Strain Distributions in Mouse Femora with 3D-Digital Speckle Pattern Interferometry

Bone is a mechanosensitive tissue that adapts its mass, architecture and mechanical properties to external loading. Appropriate mechanical loads offer an effective means to stimulate bone remodeling and prevent bone loss. A role of in situ strain in bone is considered essential in enhancement of bone formation, and establishing a quantitative relationship between 3D strain distributions and a rate of local bone formation is important. Digital speckle pattern interferometry (DSPI) can achieve whole-field, non-contacting measurements of microscopic deformation for high-resolution determination of 3D strain distributions. However, the current system does not allow us to derive accurate strain distributions because of complex surface contours inherent to biological samples. Through development of a custom-made piezoelectric loading device as well as a new DSPI-based force calibration system, we built an advanced DSPI system and integrated local contour information to deformation data. Using a mouse femur in response to a knee loading modality as a model system, we determined 3D strain distributions and discussed effectiveness and limitations of the described system.     

辐射源旋转台阶成像技术

 提出了用旋转台阶对辐射源成像的方法。对台阶响应曲线微分可得到辐射源2维出射强度沿台阶方向的1个投影，通过旋转台阶体，可得到完整的正弦图，从而可用层析重建算法重建出辐射源区的强度分布。用钨合金台阶体在1个60Co源上进行成像实验，从400个角度的台阶响应数据重建出60Co源活性区图像，表明该方法是可行的。此方法可推广到大孔径圆孔台阶体，用于测量X射线焦点等小尺寸辐射源的尺寸。     

消除数字光刻像素栅格衍射影响的研究

提出用优化成像系统设计参量的方法有效消除基于数字微镜阵列的栅格结构及其衍射对微结构成像质量的影响.模拟结果表明,当照明光源的波长为0.365 μm,系统的数值孔径为0.3左右,缩小倍率为10～20×时,栅格像的可见度在0.1以下,DMD的栅格结构对成像质量的影响大幅降低;对微透镜的成像分析发现,合理的成像系统结构参量能有效地减少栅格的影响.如果结合DMD显示图形的可编程特点,优化图形的结构或灰度,则能在像面获得较理想的曝光量分布,达到快速加工二维和三维微结构的目的.实验结果证实了优化系统参量可有效消除DMD栅格的影响.