小缺点

The active contour model is a powerful image segmentation method for representing image contour, but it has three severe difficulties. First, active contour model highly depends on its initial position. Second, active contour model lacks the ability to handle boundary concavities.

动态轮廓模型(active contour model； ACM)是一种功效强大的影像分割描绘方法，但却具有三个严重的缺点：第一、此模型非常依赖待描绘图形的起始位置，第二、此模型缺乏进入图形凹陷处的能力，第三、此模型有停滞在局部能量极小值的倾向。

After repeated studies, we found that gradient vector flow deformable model is robust to model initialization and not stagnant by local energy minimum, but still is difficult to handle boundary concavities. Thus, we develop a modified technique to overcome this drawback. This technique makes active contour further progress into concavities.

经过反覆实作，我们发现GVF较不受模型初始值之影响，而且不会因局部极小能量而停滞，然而此法仍难以进入凹陷处，因此我们也研发了一种改良技术来克服这个缺点。

Two algorithms breaks through the shortcoming of weighting value immobilizes in the traditional average filter algorithm, and it can characteristic adjust the weights from of data in the filter window, thus makes the weights from noise point is very small contributing to the output date.

两种算法突破了传统的均值算法中加权值固定不变的缺点，可以根据窗口内数据的特点来自适应的调整赋予每个采样点的权值，从而使受噪声污染的信号对输出的权值贡献非常小，而且通过对滤波因子的调整，使得对于脉冲噪声和高斯噪声均有良好的抑制作用。

In view of these shortcomings, this frequency meter has made the radical innovation in the design, its merit is: Uses the core component is the advanced monolithic special-purpose frequency measurement component -- monolithic frequency meter, the integration rate is high, the volume is small, consumes the electricity province, function, has realized the frequency meter high integration and the microminiaturization; So long as the monolithic frequency meter adds on the crystal oscillator, the measuring range choice, the frequency demonstrated and so on the very few components then constitute DC to the 10MHz miniature basic frequency measurement electric circuit; But the biggest merit was this frequency meter has realized monolithic frequency meter, frequency sampling electric circuit and the monolithic microcomputer three between hardware and the software connection completely, caused the frequency measurement measuring range the choice, the frequency data survey, the sampling as well as the code frontier transformation and the data dump possible monolithic microcomputer software programming automatically to carry on, thus has realized the frequency measurement and the sampling work intellectualizes completely, causes this system already to be possible to constitute a miniature intelligence to measure rate the instrument core electric circuit independently, also might take in the large-scale automatic control or a test system intelligent subsystem.

针对这些缺点，本频率计在设计上作了根本的革新，其优点是：所用核心器件是先进的单片专用测频器件——单片频率计，集成度高，体积小，耗电省，功能强，实现了频率计的高度集成化和微型化；单片频率计只要加上晶振、量程选择、频率显示等很少量的器件即可构成一个DC至10MHz的微型基本测频电路；而最大的优点则是本频率计完全实现了单片频率计、频率采样与单片微机三者之间的硬件与软件接口，使得测频量程的选择、频率数据的测量、采样以及编码的边境转换和数据的转储均可能过单片微机的软件编程自动进行，从而实现了测频与采样工作的完全智能化，使得本系统既可独立构成一个微型智能测率仪器的核心电路，也可作为大型自动控制或测试系统中的一个智能子系统。

Although different kinds of polymerizing methods with several mixing proportion of monomer can improve its performance, but this improvement is very limited.

但吸湿、再粘、膜不强韧是以往这类浆料的缺点，使它们上浆时只能以小比例应用。

The forecast algorithm is based on the peculiarity of 3D images i.e the deformation of the contour along the space or time axis is continuous in 3D images. The forecast algorithm analyzes the trend of the deformation of the contour in the segmented images, then prognosticates the location and the shape of the contour in the next image.

三维图像中物体的轮廓沿着空间轴或时间轴的位移和形变是连续的，该算法根据这一特性，利用前几幅图的分割结果来预测当前图像中物体轮廓的位置和形状，弥补了主动轮廓模型搜索范围小的缺点。

The forecast algorithm is based on the peculiarity of 3 imag es i.e the deformation of the contour along the space or time axis is continuous in 3 imag es . The forecast algorithm analyz es the trend of the deformation of the contour in the segmented imag es , then prognosticates the location and the shape of the contour in the next image .

三维图像中物体的轮廓沿着空间轴或时间轴的位移和形变是连续的，该算法根据这一特性，利用前几幅图的分割结果来预测当前图像中物体轮廓的位置和形状，弥补了主动轮廓模型搜索范围小的缺点。

The forecast algorithm is based on the peculiarity of 3D imag es i.e the deformation of the contour along the space or time axis is continuous in 3D imag es . The forecast algorithm analyz es the trend of the deformation of the contour in the segmented imag es , then prognosticates the location and the shape of the contour in the next image .

三维图像中物体的轮廓沿着空间轴或时间轴的位移和形变是连续的，该算法根据这一特性，利用前几幅图的分割结果来预测当前图像中物体轮廓的位置和形状，弥补了主动轮廓模型搜索范围小的缺点。

In this research, B2 gene of HCV was highly expressed as a fusion protein (Trx-B2) by cloning into pThioHisA, and induced specific anti-HCV antibody in higher titer than unmodified B2 protein after immunize mice or rabbit.The results showed antigen Trx-B2 has obvious immunogenicity and induced specific antibody,which might be able to serve as diagnosis HCV. Consequently, this PcAb of rabbit was used to coated Immulon plate, and a method of both antibodys sandwiching antigens was developed to detect 32 HCV positive serum and 32 nomal serum.Results show that ratio of positive are 87.5%and 43.75% respectively.

但缺点是该抗原所诱导的抗体滴度较低，可能是由于其分子量较小，所以本课题将人工合成的HCV复合多表位抗原基因B_2克隆到表达载体pThioHisA中，利用硫氧还蛋白融合方式对该抗原进行修饰，并在大肠杆菌中表达融合蛋白(Trx-B_2)，纯化该蛋白后免疫小鼠及家兔，结果表明用硫氧还蛋白修饰后的融合蛋白免疫小鼠和家兔后，得到的抗体滴度均高于未经修饰过的复合多表位抗原合成肽所诱导的抗体滴度，并从免疫后的兔血清中提取IgG包被酶联板，利用双抗夹心法分别检测了32份HCV阳性血清和32份正常人的血清，结果阳性率分别为87.5%和43.75%，符合率为71.88%。

The agile coefficient is defined to denote the swerving difficulty index of the robot.

这种转弯方式的优点是转弯半径能够达到很小，其缺点是转弯困难，将消耗机器人大量功率。

The split between the two groups can hardly be papered over.

这两个团体间的分歧难以掩饰。

This approach not only encourages a greater number of responses, but minimizes the likelihood of stale groupthink.

这种做法不仅鼓励了更多的反应，而且减少跟风的可能性。