基于特征值分解的图像边界与空间关系描述

针对图像识别、检索、配准中的特征提取与特征描述问题,利用加权图的邻接矩阵特征值分解方法,构造具有平移、缩放、旋转不变性的图像边界形状和空间关系描述子。理论分析与仿真实验结果表明,该算法在缩放和旋转条件下具有较强的鲁棒性和较好的类别可分离性,相比传统算法可以更好地进行图像描述,且计算量有所减小。     

出口分支语句的向量化方法

传统的向量化方法和超字并行方法依靠数据依赖关系分析确定程序中的并行性,而依赖关系分析无法处理非结构化控制流语句,现有的编译器对该类语句的向量化能力有限。为此,给出一种面向SIMD扩展体系结构的出口分支语句向量化方法,该方法针对一个向量因子内的出口分支语句,能够有效地进行自动向量化处理。测试结果表明,该方法既充分发掘了程序数据流中的并行性,又保证了控制流语义的正确性。     

一种带渐消因子的偏差配准和目标跟踪算法

在多平台多传感器跟踪系统中,提出一种带渐消因子的偏差配准和目标跟踪算法,结合配准偏差和目标状态形成扩维状态变量以构造新的偏差配准模型,在跟踪过程中引入渐消因子对突变的目标状态变量进行快速响应。仿真结果表明,该算法能使系统偏差估计迅速收敛到真实值附近,在偏差发生突变时,具有较好的自适应性,并且可以提高系统的整体跟踪精度。     

一种综合空域和时频域的数字水印算法

提出一种空域和时频域相结合的多分辨率灰度图像水印算法。该算法对原始载体图像进行离散小波分解,并对灰度水印图像先进行拉普拉斯金字塔分解再进行位平面分解。将水印图像位平面分解系数嵌入到原始图像的水平和垂直分量上,水印嵌入小波系数的位置由具有混沌特性的Arnold映射确定,从而实现水印的盲提取。实验结果证明,该算法对图像的普通处理,如噪声、滤波、剪切、JPEG压缩和旋转等,均具有较好的鲁棒性。     

LDLT分解协处理器的并行结构研究

为提高LDLT分解协处理器的性能,基于FPGA平台,研究其并行结构。分析循环片间的数据依赖关系,提出LDLT分解细粒度并行算法,并在可扩展一维阵列处理器中加以实现,利用主机、算法加速器组成单精度浮点LDLT分解协处理器的并行结构。实验结果表明,与运行在2.50 GHz Pentium微处理器上的C代码相比,该协处理器可获得32.03倍~43.25倍的性能提升。     

基于多Agent的医疗信任模型的模拟

结合直接交互信任评价和第三方信誉推荐,提出了一种基于多Agent技术的医疗信任模拟模型。使用置信因子来综合直接交互信任值和第三方信誉推荐值;避免了患者直接交互的盲目性和巨额计算,完善了对医生自我推荐的辨别能力。实验采用Swarm仿真平台模拟医疗信任模型的交互过程,结果表明该模型在确定信任关系中相对IT模型要快,比第三方信誉推荐模型相对要可靠。     

应用改进的C-V模型分割牛肉眼肌图像

牛肉眼肌图像的分割是牛肉自动分级的重要步骤。基于水平集的C-V模型采用两个具有不同均值的分块连续区域分割图像,在迭代过程中,为了约束水平集保持为符号距离函数,必须对其进行重新初始化,大大增加了计算量,降低了曲线演化速度。为此,通过加入边缘检测函数和惩罚项因子,提出了改进的C-V水平集模型,并用于牛肉图像分割,同时与模糊C均值聚类、阈值分割法进行了对比。结果表明:采用改进的C-V模型对牛肉图像进行分割,准确提取了眼肌边缘,分割出了脂肪和肌肉区域,准确率高;曲线演化过程中无需对水平集函数符号初始化,加快了曲线演化速度。     

Computing integrals over polynomially defined regions and their boundaries in 2 and 3 dimensions

We use the cylindrical algebraic decomposition algorithms implemented in Mathematica to produce procedures to analytically compute integrals over polynomially defined regions and their boundaries in two and three dimensions. Using these results, we can implement the divergence theorem in three dimensions or the Green's theorems in two dimensions. These theorems are of central importance in the applications of multidimensional integration. They also provide a strong correctness test for the implementation of our results in a computer algebra system. The resulting software can solve many of the two and some of the three dimensional integration problems in vector calculus textbooks. The three dimensional results are being extended. The results in this paper are being included in an automated student assistant for vector calculus.     

The complex local mean decomposition

The local mean decomposition (LMD) has been recently developed for the analysis of time series which have nonlinearity and nonstationarity. The smoothed local mean of the LMD surpasses the cubic spline method used by the empirical mode decomposition (EMD) to extract amplitude and frequency modulated components. To process complex-valued data, we propose complex LMD, a natural and generic extension to the complex domain of the original LMD algorithm. It is shown that complex LMD extracts the frequency modulated rotation and envelope components. Simulations on both artificial and real-world complex-valued signals support the analysis.     

Structure damage diagnosis using neural network and feature fusion

A structure damage diagnosis method combining the wavelet packet decomposition, multi-sensor feature fusion theory and neural network pattern classification was presented. Firstly, vibration signals gathered from sensors were decomposed using orthogonal wavelet. Secondly, the relative energy of decomposed frequency band was calculated. Thirdly, the input feature vectors of neural network classifier were built by fusing wavelet packet relative energy distribution of these sensors. Finally, with the trained classifier, damage diagnosis and assessment was realized. The result indicates that, a much more precise and reliable diagnosis information is obtained and the diagnosis accuracy is improved as well.