特征方向

The new method is a combination of characteristic approximation to handle the convection part, to ensure the high stability of the method in approximating the sharp fronts and reduce the numerical diffusion, a smaller time truncation is gained at the same time, and a mixed finite elementspatial approximation to deal with the diffusion part, the sealer unknown and the adjoint vector function are approximated optimally and simultaneously.

此方法即为对方程的对流项沿流体流动的方向即特征方向进行离散，从而保证格式在流动锋线前沿逼近的高稳定性，消除了数值弥散现象，并得到了较小的时间截断误差；另一方面，对方程的扩散项采用混合元离散，可同时高精度逼近未知函数及其伴随向量函数，理论分析表明，此方法是稳定的，具有最优的L~2逼近精度。

The characteristic approximation is used to handle the convection part along the direc-tion of fluid namely characteristic direction to ensure the high stability of the method in approximating the sharp fronts and reduce the numerical diffusion; The mixed finite element spatial approximation is employed to deal with diffusion part and approximate the scalar unknown and the adjoint vector function optimally and simultaneously; In order to preserve the integral conservation of the method, we introduce the modified characteristic method.

该方法对方程的对流部分沿流体流动的方向即特征方向离散以保证格式在流动的锋线前沿逼近的高稳定性，消除数值弥散现象；对方程的扩散部分采用最低次混合有限元方法离散、同时以高精度逼近未知函数及未知函数的梯度；为保证方法的整体守恒性，在格式中引入修正项。

A new anisotropic diffusion filtering method which is based on this framework is proposed in this paper.

所以，基于图像特征方向建立的各向异性扩散滤波方法更能达到我们预期的效果，该设计方法是有效的。

Douglas and Russell first presented the idea of characteristics in[13],in whichthe diffusion term is included in the derivative of characteristic direction of the solu-tion and the backward difference quotient along the time axis was replaced by it alongcharacteristic direction.

Douglas，Russell[13]最先提出了特征法的思想，其主要特点是将模型方程的对流项隐含在解沿特征方向的导数之中，以沿特征方向的向后差商逼近取代通常方法中沿时间方向的向后差商逼近。

These algorithms firstly get the characteristic of line from the normal part, and then track the line under the guidance of the characteristic direction, so that it can track through intersection or overlapping parts to recognize an entire line in one step.

这些方法在线条的无干扰部分搜索其特征，然后在特征方向的指导下进行跟踪，可以避免线条上相交、粘连部分的干扰，完整地识别线条。

For the bottom-up attention process, areas of interest that deserve visual attention are formed by a sequentially extracting of feature map, conspicuity map, and interesting map. Firstly, based on Treisman's feature integration theory and the characteristics of orientation and frequency selective nerve cells in biological cortex, we propose a new 2D optimal quadrature filter in the meaning of minimum energy loss to extract intensity, edge, orientation, symmetry, curvature and comer features to form the feature maps. Then based on the local competent and global cooperative perception character of biological vision, we design a kind of competent/cooperative filter and use a set of such filters to extract conspicuous locations that will most likely to cause visual attention from feature maps to form conspicuity maps.

我们根据Treisman〓的特征整合理论，结合生物视觉皮层的方向、频率敏感性神经元，提出采用一组调制在不同尺度和方向上的新的最佳正交特征滤波器并行地提取图象的亮度、边缘、方向、对称性、曲率和角度等特征作为注意线索，构成特征图；在此基础上，基于生物视觉的局部竞争和全局协作特性，本文设计了一种竞争协作滤波器用于从特征图中提取可能引起视觉注意的显著位置构成显著图，各显著图进一步通过本文提出的融合算法和非线性松弛迭代算法生成自底向上兴趣图。

A set of complex Morlet wavelet are applied on projection slice of each direction to decompose each projection into several frequency channels,the average and variance extracted are computed in each frequency channel,and then linear regression model is employed to computer realationship feature between frequency channels.1-D DFT is applied to features and the amplitudes of Fourier coefficient are selected as features,then the extracted features are rotation invariant.

该算法是将一定大小的图像进行二维傅里叶变换；其次在变换后的图像中央选择一个圆盘区域，并在方向［0°，180°］内进行等间隔角度频率抽样，实现方向分解，使用一组复Morlet小波对每个方向上的映射切片进行小波变换，从而实现多通道频率分解；在各个频率通道中计算均值和方差作为特征，并利用线性回归模型计算频率通道之间的关系特征；将特征沿方向进行一维傅里叶变换并取其幅值，从而得到旋转不变性特征。

First, the proposed scheme uses contourlet transform to extract multidirectional and multiscale texture information, and then obtains the feature points invariant to the affine transformation from the low and middle directional subbands by Harris-Affine detector. Second, the feature regions are adaptively computed by the feature scale of the local structure and normalized by U transform. The watermarking is embedded adaptively in those characteristic regions which are normalized. Finally, by vector quantization, several copies of the watermark are embedded into the no overlapped local feature regions in different directional subbands.

首先，用Contourlet变换提取出多尺度、多方向的纹理信息；再用Harris-Affine检测算子从变换域的中、低频方向子带中提取出仿射不变特征点，结合自适应局部结构的特征尺度确定特征区域，并用U变换对其归一化处理，水印就自适应地嵌入到归一化后的区域中；水印嵌入采纳矢量量化的策略，将水印信息重复嵌入到不同方向子带、多个不相交的局部仿射不变特征区域。

In order to describe quantitatively their relative contributions, the kinematic vorticity number is introduced and simply defined as cos v, where v is the angle between two eigenvectors containing the shear directions in the principal deformation plane （XZ-plane or ac-plane）.

为了定量说明两者间的相对贡献，提出了运动学涡度这一物理量，并简单地定义为COSv.v是主变形面内两特征方向间的夹角。

Based on the Roe′s approximate Riemann solver, this paper presents the flux balance Godunov scheme for the shallow water equation with source terms.

以Roe的近似Riemann解为基础，将源项按特征方向进行特征分解，建立了带源项浅水方程的通量平衡Go dunov求解格式。

He is the most remarkable man that I have ever known.

他是我知道的人中最值得注意的人。

Increased appearance of lymphocele was noticed in patients with diabetic nephropathy, congenital malformations of the urinary tract, and inflammatory diseases, including glomerulopathy and extraglomerular ones, after high-voltage radiotherapy and after removal of the renal graft.

经过高压放射疗法和切除移植肾后，在糖尿病肾病患者、泌尿系统先天性畸形患者和炎症患者身上发现淋巴囊肿有所增加，包括肾小球病和球外疾病等。