在算法上

The key to feature extraction is to select the optimal feature combination from all the features. In this study, the optimal color feature combination,{/G,(R+4G+B)/6, H variance, S variance}, is selected using genetic neural network at first. Then the main texture features including the mean, variance, energy and entropy of the cooccurrence matrix, the secondary texture features including the maximum frequency number, absolute uniform, inertia and importance of the co-occurrence matrix are extracted.

在此基础上，提出适合多阈值和小目标图像分割的免疫遗传图像分割方法，该方法借鉴了生物免疫系统自调节、抗原识别和记忆功能，并结合了遗传算法对图像进行自适应区域的阈值分割，改善了遗传算法优化复杂问题时存在收敛速度慢和易陷于局部最优的不足，克服了传统阈值分割方法对目标大小的限制，是一种新型的有效的图像分割方法。

Based on the optimal data partitioning algorithm, only the accumulated error from the past frames is taken into account, and the complexity and performance of the causal optimal data partitioning algorithm are studied.

在最优数据划分算法的基础上，仅考虑过去帧的积累误差，不考虑误差对将来帧的影响，探讨了因果最优数据划分算法的复杂度及性能。

A prototype system is developed finally with ACIS 7.0 and Visual C++ 6.0, and some samples are used to proved the algorithms involved in this paper.

根据本文提出的算法，在几何造型系统ACIS 7.0和VC++6.0平台上开发了组合夹具夹紧方案自动规划的原型系统，该系统通过对几个实例测试，验证了本文提出的算法。

In order to reduce calculation in the process of vertical scale identification of torpedo acoustic image,we present an improved decoherent multiple signal classification method based on quadrature complex sampling which suits to estimating the vertical highlights of submarine with the characteristics of near distance and coherent signals .

在鱼雷声成像垂直尺度识别中，根据潜艇垂直方向上目标亮点相距较近，且为相干信号源的特点，同时为了减小运算量而易于工程实现，提出了一种基于正交采样的高分辨DOA估计算法，即正交解相干MUSIC算法。

A novel direction of arrival and polarization estimation of electromagnetic vector sensors mounted on the airframe is proposed based on biquaternion. The signal processing model for EM-sensor mounted on the airframe is made by biquaternion and then the eigenvalue decomposition of spectral matrix is obtained by means of the EVD of its quaternion adjoint matrix. The memory requirement for data is reduced, and more accurate estimation result is gotten by deducing the stronger orthogonality constraint between the sin- gal and noise subspaces on the field of biquaternion. The performance of the new algorithm is confirmed through numerical examples.

针对机载电磁矢量传感器阵列DOA和极化参数估计问题，提出了一种基于复四元数估计方法，该算法利用四元数建立机载电磁矢量传感器阵列信号处理模型，然后利用四元数联合矩阵的特征分解得到阵列数据相关矩阵的特征分解，一方面使得计算过程中数据的贮存量大大减少，另一方面通过推导得到信号子空间和噪声子空间在四元数域上的正交性从而使DOA和极化参数佑计的精度更高，仿真证实了本算法的有效性。

Applying this algorithm over GF(28) the algebraic expression of Rijndael S box is obtained.

在GF(2~8)上应用此算法，得到了Rijndael S盒的代数表达式，且任意改变Rijndael S盒的仿射变换，都能应用该算法求出其代数表达式。

This research uses AutoCAD as model platform,which is further developed by Visual Basic language ,and demonstrates that the two ameliorative algor.

本研究以Au-toCAD作为建模平台，应用 VB语言对其开发，通过实验所得的结论证明所提出的两种改进算法均能较传统算法在运行效率上有极大的提高。

This algorithm is more efficient in solving register allocation problem than clique partitioning algorithm.

该算法在解决寄存器分配问题上比常采用的团划分算法更有效。

We use O(n2) processors in the study. By combining the characteristics of PARES and improving the methods of [14][15], we are able to derive constant-time algorithms for this problem.

关於环弧图上的支配问题，以前尚未有人在常数时间内解决，即使是在理想的PRAM模型上也是如此，在本论文中，我们改良Yu[14][15]中提到的相关演算法，然后再自行发展出一套理论，并且将之推广至可重组态汇流排之处理器阵列模型上。

This paper introduces the background knowledge of heterogeneous data integration, as well as the knowledge of data integration. On the basis of the above, The author develops an ontology-based Heterogeneous data Integration prototype system, expounds the main framework and essiontial technology. Then the paper introduces its work in detail: it explains the very significance of using ontology as an comon semantic description and an global query model ,and illustrates the reason fo adopting LAV method ; then it builds the formal representation of the mapping between ontology and data resources, applies OWL to describe the ontology and the mapping ; defines a query language named by DIQL to query on our integration system , and supply the detailed arithmetic of reolving partial binding of LAV method、locating of data sources and query rewriting,etc. Finally,the paper demostrates and proves the result of research by examples .

本文介绍了目前异构数据集成的研究现状以及数据集成和本体的相关背景知识，并在此基础上设计开发了一个基于本体的异构数据LAV集成原型系统，同时对原型系统的总体框架、关键实现技术分别做了讨论：首先阐述了将本体作为公共语义描述和全局查询模型的意义，给出了采用LAV集成方案的理由，然后规定了本体和数据源间映射的形式化表示方式，采用能结合描述逻辑和框架系统的优点、适合于半结构化环境的OWL表示本体和映射；其次定义了一种针对集成系统的查询语言DIQL，并在此基础上给出解决LAV部分绑定问题、数据源定位以及查询重写的算法；最后在原型系统的基础上，通过实例验证本文所提方法的可行性。

For a big chunk of credit-card losses; the number of filings (and thus charge-off rates) would be rising again, whether

年美国个人破产法的一个改动使得破产登记急速下降，而后引起了信用卡大规模的亏损。

Eph. 4:23 And that you be renewed in the spirit of your mind

弗四23 而在你们心思的灵里得以更新