体上拟伸缩

设K是一个给定的体,用GLn(K)表示体K上的n级一般线性群,用resA表示矩阵A的剩余数.在文献[3]基础上,对拟中心矩阵的概念进行重新定义,并得到相应结果.同时,引入拟伸缩矩阵的概念,并对它进行了刻画.

基于全局和区域可伸缩拟合局部熵活动轮廓模型的超声图像分割

针对区域可伸缩拟合局部熵(region-scalable fitting based on local entropy,RSF_LE)模型图像分割效率低的问题,本研究提出一种改进的RSF_LE模型。定义带有加权局部灰度拟合项以及辅助的加权全局灰度拟合项的能量泛函,其中加权局部灰度拟合项负责对目标边界附近的轮廓进行诱导,使其靠近目标物边界,加权全局灰度拟合项利用图像的全局信息来引导远离目标的轮廓向目标靠拢,该方法可以克服传统的RSF_LE模型分割算法效率低下的问题,并提高了该方法的鲁棒性。

Modeling and simulation of a high-frequency micro-pump based on giant magnetostrictive material (GMM)

A new type of high-frequency micro-pump was designed, in which GMA (Giant Magnetostrictive Actuator) was employed to replace the traditional motor drive, owing to its significant characteristics of fast response, high accuracy, easily miniaturized and so on. Both the mathematic and simulation models of the micro-pump were built.A set of raw data was used for simulation studies.The results show that the micro-pump based on GMA has achieved the features of high-frequency response and high accuracy, theoretically displaying the performance merits of a giant magnetostrictive material (GMM) high-frequency micro-pump.

联合小波变换和RSF模型的CT图像分割方法

为解决工业计算机层析成像(CT)图像的伪影和弱边缘问题,提出一种基于小波变换的图像区域可伸缩拟合能量最小化分割方法,实现图像边缘的精确定位,从而提高图像测量精度。首先,采用小波变换对图像进行预处理,降低金属伪影。然后,采用所提方法精确分割图像,提高感兴趣区域边缘的定位精度。实际数据测量结果表明,所提方法可有效降低图像弱边缘的影响,测量相对误差低于0.7%,相较Chan-Vese算法,测量精度提高了1.4倍,满足实际测量需求。