基于对地观测激光回波特征的目标特性研究

本文研究了对地观测激光回波特征与地面目标特征之间的关系模型，分析了分类地面目标的激光回波特征，提出一种基于查表法的目标特性计算方法并给出了仿真结果。     

小波提升算法及其在图像处理中的应用

小波提升算法是一种新的双正交小波构造方法，通过预测算子，确定高频信息，并初步确定低频信息，然后通过更新算子，对初步确定的低频信息进行修正，从而确定低频信息。它在空域对信号进行变换，完成了对信号频域的分析。在图像处理中，基于离散小波变换的提升算法比传统的卷积算法运算简单，实时性好，易于实现，因而被新一代图像压缩标准JPEG2000所采用。文中简要介绍了小波提升算法的原理，分析了其特点，并介绍了JPEG2000标准中采用的W5／3、D9／7两种小波的提升格式和实现算法。     

Blind Channel Identification Based on Noisy Observation by Stochastic Approximation Method

A stochastic approximation algorithm for estimating multichannel coefficients is proposed, and the estimate is proved to converge to the true parameters a.s. up-to a constant scaling factor. The estimate is updated after receiving each new observation, so the output data need not be collected in advance. The input signal is allowed to be dependent and the observation is allowed to be corrupted by noise, but no noise statistics are used in the estimation algorithm.     

红外化学遥感数字信号处理算法的研究进展

介绍了几种为远距离红外蒸汽信号监测开发的分类器设计，频域数字滤波，时域数字滤波等数字信号处理技术。对线性分类器，分段线性分类器和神经网络分类器等在红外化学遥感系统中的应用做了初步的探讨。对背景扣除技术，频域数字滤波技术，数字滤波的切趾方法等做了简要的描述。阐述了时域滤波技术的基本原理及实际应用。展望了红外化学遥感数字信号处理算法的研究前景。     

关于Hopf丛S3的几个论题

本文证明了Hopf主纤维丛S^3的几个相关的命题，指出底流形S^3上的Laplace算子在主丛上的提升是主丛S^3上的Laplace算子，以及主丛的嵌入截面具有不动点性质等。     

Surface morphologies of anthracene single crystals grown from vapor phase

The morphologies and lattice structures of anthracene single crystals grown from the vapor phase were investigated using optical microscopy, phase contrast microscopy, atomic force microscopy (AFM), and X-ray diffraction analysis. Common morphologies with hexagonal large planes were observed irrespective of crystal size. The observation of certain surface morphologies with a phase contrast microscopy revealed that the spiral steps originated from screw dislocations present on the (0 0 1) planes. Moreover, the center and edge of the (0 0 1) planes had large curvatures, similar to hills. Resultantly, quarter-monolayer (ML) steps were observed on the large and flat planes between both hills.     

基于移动Agent技术的医学图像检索方法

针对现有医学图像检索方法中，检索算法速度较慢、准确性较低以及远程检索时占用大量带宽等问题，设计并实现了一种基于移动Agent技术的医学图像检索方法，重点讨论了该方法实现的关键技术，包括提升框架下的M带整数小波变换和提取多通道纹理特征的检索算法．模拟实验结果表明，该方法实时性较强，占用网络带宽较少，采用的算法运算速度提高近10倍、运算复杂度降低约50％，提取的纹理特征能充分表达图像内容，较好地满足了医学图像的检索要求，实现了对医学图像库快速较准确的检索．     

利用弧形光栅尺测量大型转轴扭矩的原理研究

利用弧形光栅尺的旋转微位移测量功能,组建一组测量体系,用于测量巨型转轴的扭矩。采用四组相同的弧形光栅尺同时测量两截面的相对转角,通过多点测量实现误差均化,得出了消除转轴直径对测量影响的公式,使仪器的环境适应能力得到了极大地提高。根据不同的工作情况,提出了相应的动态测量扭矩的计算方程。实现了大型机械设备主传动的动力监测,为设备自动化、智能化奠定了基础。     

In situ observation of NaBi(WO4)2 dendrite growth and the study on microstructure of solidification

The dendrite growth process of transparent NaBi(WO₄)₂ with small prandtl and high melting point was studied by using the in-situ observation system. According to the dynamic images and detailed information, there are two kinds of restriction effect on the dendrite growth, the competition between arms and branches and the convection in the melt. The dendrite growth rate was time dependent, and the rate of arm growth reached the maximum 5.8 mm/s in the diffusive-advective region and rapidly decreased in the diffusive-convective region. The growth rate of branch had the same change trends as the arm’s. Based on the EPMA-EDS data of solidification structure of quenched NaBi(WO₄)₂ melt, it was found that there were component differences from stoichiometric concentration in the melt near the interface during the growth process. Supported by the National Natural Science Foundation of China (Grant No. 50331040) and the Innovation Funds from Shanghai Institute of Ceramics, Chinese Academy of Sciences (Grant No. SCX0623)     

Sharp precision in Hensel lifting for bivariate polynomial factorization

Popularized by Zassenhaus in the seventies, several algorithms for factoring polynomials use a so-called lifting and recombination scheme. Concerning bivariate polynomials, we present a new algorithm for the recombination stage that requires a lifting up to precision twice the total degree of the polynomial to be factored. Its cost is dominated by the computation of reduced echelon solution bases of linear systems. We show that our bound on precision is asymptotically optimal.