The Strong Property of Morphological Connected Alternated Filters

This paper studies connectivity aspects that arise in image operators that process connected components. The focus is on morphological image analysis (i.e., on increasing image operators) and, in particular, on a robustness property satisfied by certain morphological filters that is denominated the strong property. The behavior of alternated compositions of openings and closings is investigated under certain assumptions, particularly connectedness and a connected component preserving condition. It is shown that these conditions cannot in general guarantee the strong property of certain connected alternated filters because of issues related to the locality of the filters. As treated in the paper, there have been a series of misunderstandings in the literature concerning this topic, and it is important to clarify them. The root cause of those problems is discussed, and a solution is indicated. The class of connected openings and closings used to build connected alternated filters should therefore be defined to avoid such situations, since the strong property of alternated filters should be a distinctive characteristic of this class.

Topological and Variational Properties of a Model for the Reconstruction of Three-Dimensional Transparent Images with Self-Occlusions

We introduce and study a two-dimensional variational model for the reconstruction of a smooth generic solid shape E, which may handle the self-occlusions and that can be considered as an improvement of the 2.1D sketch of Nitzberg and Mumford (Proceedings of the Third International Conference on Computer Vision, Osaka, 1990). We characterize from the topological viewpoint the apparent contour of E, namely, we characterize those planar graphs that are apparent contours of some shape E. This is the classical problem of recovering a three-dimensional layered shape from its apparent contour, which is of interest in theoretical computer vision. We make use of the so-called Huffman labeling (Machine Intelligence, vol. 6, Am. Elsevier, New York, 1971), see also the papers of Williams (Ph.D. Dissertation, 1994 and Int. J. Comput. Vis. 23:93–108, 1997) and the paper of Karpenko and Hughes (Preprint, 2006) for related results. Moreover, we show that if E and F are two shapes having the same apparent contour, then E and F differ by a global homeomorphism which is strictly increasing on each fiber along the direction of the eye of the observer. These two topological theorems allow to find the domain of the functional ℱ describing the model. Compactness, semicontinuity and relaxation properties of ℱ are then studied, as well as connections of our model with the problem of completion of hidden contours.

Decentralised control of chromatographic simulated moving bed processes based on wave front reconstruction

The behaviour of chromatographic simulated moving bed processes is described by the movement of concentration profiles through a circle of separation columns. A closed-loop control manipulates the profiles in order to meet demands concerning specified product purity and disturbance attenuation. If steep wave fronts of the concentration profiles occur, the controlled variables undergo fast changes in case of a transient of the process. In this case, a reconstruction of the wave fronts is necessary for a successful control.A simple and effective decentralised controller structure is proposed based on cascaded discrete-time PI controllers. On-line product purity measurements and the reconstructed wave fronts are used for control purposes. Two kinds of process models are used: a rigorous model for dynamic simulations, and strongly simplified plant models for the design of the wave front reconstruction and the controller. The latter models are identified based on experimental step tests with the reference plant and numerical simulations. The performance of the control system is evaluated by numerical simulations.     

基于NMF图像重构的人脸识别

由传统的人脸识别方法产生的人脸特征子空间通常是由人脸库中所有训练样本产生的一个通用子空间,该空间更多地包含了所有人脸样本的共性特征,而忽略了个性特征。该文提出一种基于NMF图像重构的方法,以单个人的训练样本集获取其人脸特征子空间,将识别图像向每一个特征子空间中进行映射及重构,并以重构图像的误差作为判据实现人脸识别。在ORL标准人脸库进行的计算机仿真证实了该方法的有效性。     

基于改进加权一阶局域预测模型的负荷预测方法研究

针对混沌时间序列预测中用加权一阶局域法单步预测模型进行多步预测时计算量大,且存在误差累积效应的不足,本文提出了基于相空间重构技术的加权一阶局域法多步预测模型,通过实际算例分析显示,本文方法较改进前有较好的适应能力和预测精度。     

基于主动形状模型的双目识别系统的设计与实现

给出了一种基于统计形状模型的双目识别系统实现方法.该系统根据识别目标在场景中的姿态有一定规律性的特点,用先验的样本信息指导识别的过程,并采用径向对齐约束的方法对摄像头内外参数进行标定,利用图像识别结果中特征点的顺序性完成特征匹配,进而获得目标的三维信息.     

模拟退火遗传算法的轮廓线拼接研究

基于断层数据的三维重构是三维数据场可视化技术的主要研究内容,重构物体的三维模型也是分析、仿真的前提.应用模拟退火遗传算法设计和实现了三维物体表面重建算法.通过在遗传算法中融人退火处理操作,提高了种群的多样性,避免了遗传算法中存在的早熟收敛问题,有效地增强了算法的全局寻优能力.此外,提出了一种基于边的最小交叉多边形算法,提高了交叉操作效率.根据轮廓线拼接问题的特点,设计了针对性强的交叉、变异操作.实验结果表明,算法有效地提高了传统的全局法轮廓线拼接算法的效率.     

空间自适应正则化超分辨率图像重建

超分辨率图像重建是一个病态问题,在重建过程中需要正则化处理,而正则化重建会引入正则化误差及重建过程中由于病态性而引入的噪声放大误差,且这两类误差均和图像的空间局部特性有关.提出根据图像的局部空间统计特性自适应控制超分辨率图像正则化重建算法,采用图像局部统计方差来区分图像棱边区域及平滑区域,在图像的棱边区域加强图像的约束重建,而在图像的平滑区域加强正则化.实验表明该算法能有效地减小重建误差,算法的信噪比得益优于传统的正则化重建算法及总变分模型重建算法,并且对正则化参数的选择具有一定的鲁棒性.     

基于微包协议的三模冗余容错计算机无缝重构算法

提出了一种基于微包传输协议的系统重构算法,制定了系统重构命令协议,设计了关键数据区的一致性管理算法.不需增加额外的硬件资源和系统开销,利用三机之间的数据交换通道,即可实现三模冗余容错计算机的无缝重构,保证系统在故障恢复重构时工作的连续性.同时,完善的重构命令协议使得重构软件的开发可以采用面向对象思想构成标准化的中间件,具有很强的扩展能力,为构建标准化嵌入式三模冗余容错计算机内核奠定了良好的技术基础.     

一种快速的基于球体混合重建的碰撞检测算法

为了解决虚拟特征和虚拟环境中虚拟物体的相交问题,必须对实时的虚拟应用环境实施快速的碰撞检测.提出了基于球体混合重建的碰撞检测算法,对球体混合表面提出了一个子线性时间复杂性的重建过程;利用四元组构造变换范围,扩展了从线性到球形混合的构造变换方法;采用循环划分的方法完成了由球体混合到线性混合的分解过程.尽管这个方法的实现过程较线性的方法更难些,但算法的执行过程所需时间却与线性方法相差无几,其复杂性也与线性结构的相同.