基于结构张量的自适应CTV彩色图像恢复模型

讨论一种基于非线性扩散方程的彩色图像去噪方法。在图像去噪的3个基本要求的基础上,总结出调和项模型和彩色总变差去噪模型中的不足,利用图像的局部信息构造函数使得模型在接近图像边缘处各向异性平滑并保持边界。在平坦区域各向同性平滑,防止阶梯效应的产生,并利用角点信息保持了角点形状。实验结果表明,所建模型能够较好地保持图像中目标的几何结构,同时具有良好的去噪能力。     

基于小波估计和直方图特征的LSB匹配分析

针对隐写分析中的难点——空域LSB匹配隐写进行检测和分析,描述LSB匹配加性隐写的特点,将匹配隐写建模为图像受到一定强度的脉冲噪声干扰,采用小波变换对退化的图像进行恢复作为载体的估计。对检测图像和恢复图像提取多个直方图特征比值作为特征向量,利用支持向量机对500幅高质量未压缩的自然图像组成的载体、载密图像库进行检测,结果证明该算法在低嵌入率下可获得较好的检测效果。     

基于NMF分组策略的人脸识别

提出一种运用非负矩阵分解（NMF）分组策略进行人脸识别的方法。将训练图像分组,分别对每组图像作NMF,获取每组图像的基图像构成的非负特征子空间,将训练图像和测试图像分别向各个特征子空间进行投影,将每组图像提取出的特征系数混合,根据最近邻原则进行识别。基于ORL人脸数据库上的实验证明了该方法的有效性。     

基于有监督直接局部保持投影的人脸识别

提出一种用于图像识别的有监督直接局部保持投影算法,该算法结合样本类别信息,通过同时对角化的方法求解局部保持投影问题,避免矩阵的奇异性。在ORL人脸库上的测试结果表明,该算法的识别率高于PCA, PCA+LPP等方法。     

混沌敏感性在图像加密中的2种应用

混沌系统的结构参数敏感性、初值敏感性和以同一分布遍历各态的特性较好地对应密码系统的一些基本特性。分析结构参数和初始值变化对混沌序列敏感性影响,得到混沌序列敏感性效应是一个积累过程。将混沌序列按照敏感性划分为混沌敏感性前后期序列并给出界定原则。探究混沌序列敏感性在图像加密时的2种不同应用。     

基于视频对象的区域分割及其应用

为适应MPEG-4基于对象的应用,提出一种基于视频对象的区域分割算法。在视频对象的指导下,利用均值漂移进行分割,性能比传统均值漂移算法在计算时间、占用内存等方面均有提高,适用于运算速度要求高和处理器资源有限的情况。将分割结果应用于区域指导的图像插值,能保证较快的处理速度和兴趣区域较好的图像质量。实验结果证明了该方法的有效性。     

基于CPSO的二维Otsu图像分割法

二维Otsu方法同时考虑了图像的灰度信息和像素问的空间邻域信息,图像分割效果好但算法计算量较大。针对上述情况,提出一种基于混沌粒子群优化算法（CPSO）的策略,将其用于二维Otsu方法中,并与标准粒子群优化算法（SPSO）进行仿真实验对比。实验结果表明,该方法可以提高分割速度,克服SPSO的缺点,图像分割结果较理想。     

Self-consistent gradient flow for shape optimization

We present a model for image segmentation and describe a gradient-descent method for level-set based shape optimization. It is commonly known that gradient-descent methods converge slowly due to zig–zag movement. This can also be observed for our problem, especially when sharp edges are present in the image. We interpret this in our specific context to gain a better understanding of the involved difficulties. One way to overcome slow convergence is the use of second-order methods. For our situation, they require derivatives of the potentially noisy image data and are thus undesirable. Hence, we propose a new method that can be interpreted as a self-consistent gradient flow and does not need any derivatives of the image data. It works very well in practice and leads to a far more efficient optimization algorithm. A related idea can also be used to describe the mean-curvature flow of a mean-convex surface. For this, we formulate a mean-curvature Eikonal equation, which allows a numerical propagation of the mean-curvature flow of a surface without explicit time stepping.     

A New Medical Image Enhancement Based on Human Visual Characteristics

Study of image enhancement shows that the quality of image heavily relies on human visual system. In this paper, we apply this fact effectively to design a new image enhancement method for medical images that improves the detail regions. First, the eye region of interest （ROI） is segmented; then the Un-sharp Masking （USM） is used to enhance the detail regions. Experiments show that the proposed method can effectively improve the accuracy of medical image enhancement and has a significant effect.     

Finite-time weighted average consensus with respect to a monotonic function and its application

The consensus state is an important and fundamental quantity for consensus problems of multi-agent systems, which indicates where all the dynamical agents reach. In this paper, weighted average consensus with respect to a monotonic function, which means that the trajectories of the monotonic function along the state of each agent reach the weighted average of their initial values, is studied for a group of kinematic agents with time-varying topology. By constructing a continuous nonlinear distributed protocol, such a consensus problem can be solved in finite time even though the time-varying topology involves unconnected graphs. Then the distributed protocol is employed to compute the maximum-likelihood estimation of unknown parameters over sensor networks. Compared with the existing results, the estimate scheme proposed here may reduce the costs of data communication, storage memory, book-keeping and computational overheads.