基于粒子滤波的人脸图像超分辨率重建方法

将人脸图像超分辨率重建描述为人脸混合模型的纹理和位置参数的贝叶斯概率估计问题,将超分辨率重建的图像配准和像素融合这两个过程置于统一的概率估计框架下,并利用基于粒子滤波的参数估计算法,同时估计纹理和位置参数,从而实现人脸图像的超分辨率重建.包含灰度和位置两种先验信息的人脸混合模型,同时用于超分辨率重建的两个过程中,提高了图像配准精度和重建算法的性能,避免了通常方法在获得准确鲁棒的运动场估计时需要清晰的高分辨图像,而获得清晰的高分辨图像时又需要准确鲁棒运动场估计的困境.正面人脸合成序列图像实验结果表明,该方法获得的重建结果较为理想.     

基于樽海鞘群体优化非负矩阵分解的高光谱图像解混算法

针对鲁棒非负矩阵分解应用于高光谱图像处理时,存在对初始值的敏感性,求解目标函数时易陷入局部最优的缺点,提出基于樽海鞘群体优化鲁棒非负矩阵分解的高光谱图像解混算法.该算法基于鲁棒线性混合模型,在RNMF框架下,采用樽海鞘群体算法取代乘法迭代策略,以增强算法全局搜索能力,在约束空间内随机搜索满足目标函数的全局最优解,可有效地完成非线性高光谱图像解混.仿真数据与真实遥感数据实验结果表明,本文算法在处理高光谱图像时,能够有效地避免RNMF算法易陷入局部最优解的局限性,具有更好的解混性能.     

鲁棒卡尔曼滤波下的图像雅可比矩阵带时延补偿的估计

传统的图像雅可比矩阵估计的方法没有考虑时延因素,因此具有较大的估计误差.为补偿时延带来的影响,提出一种鲁棒卡尔曼滤波的方法,实现时延情况下当前时刻特征点在图像空间中位置和速度的估计,进而得到时延情况下较为准确的图像雅可比矩阵的估值.具体说,特征点在图像空间中当前时刻位置和速度是首先用卡尔曼滤波的方法估计的,但观测噪声的描述却采用了马尔科夫链模型,由此产生了过程噪声和观测噪声的互相关,传统卡尔曼滤波受限.为此,我们引入滤波修正向量并重新定义过程方程及观测方程,结合卡尔曼滤波中噪声的数学特性,得到滤波修正向量消除互相关性,从而构建出鲁棒卡尔曼滤波模型;其次,针对鲁棒卡尔曼滤波模型中存在的无法获得时延期间的观测向量的问题,提出利用多项式拟合出这部分观测向量,该多项式的选取综合考虑了特征点的位置、速度、加速度、加速度的变化率对于特征点轨迹的影响,与实际情况相符;最后,由预测出的当前时刻特征点在图像空间中的位置和速度,实现时延情况下图像雅可比矩阵较为准确的估计.仿真和实验结果验证了本文方法的可行性和优越性.     

计算机视觉中强鲁棒性的遗传一致性估计

用于估计计算机视觉模型的传统鲁棒算法均存在估计精度和稳定性不高等问题。为此,结合遗传算法的全局最优性及几何模型估计的特殊性,提出一种强鲁棒性的遗传一致性估计算法,以估计各种误差和错误概率下的计算机视觉几何模型。仿真实验结果表明,相比于RANSAC、MAPSAC、MLESAC等鲁棒算法,该算法在估计精度和鲁棒性方面性能更优。     

基于引力模型的三维运动点集的非精确匹配

本文提出了一种基于引力模型(attractive model)的非精确匹配算法,应用于三维空间运动点集的对应点匹配问题.根据引力模型,我们将匹配和运动估计问题转化为一个代价函数的全局优化问题,实现了无对应点的运动估计和总体匹配.这种算法是一个鲁棒(robust)估计和匹配方法,可以处理包含非匹配点对的三维运动点集.大量计算机模拟实验结果充分证明了算法的鲁棒性和有效性.     

自适应最小误差阈值分割算法

对二维最小误差法进行三维推广, 并结合三维直方图重建和降维思想提出了一种鲁 棒的最小误差阈值分割算法. 但该方法为全局算法, 仅适用于分割均匀光照图像. 为 提高其自适应性, 本文采用Water flow模型对非均匀光照图像进行背景估计, 以此获 得原始图像与背景图像的差值图像, 达到降低非均匀光照对图像分割造成干扰的目的. 为进 一步提高分割性能, 本文对差值图像采用γ 矫正进行增强, 然后采用鲁棒最小误差 法进行全局分割, 从而完成目标提取. 最后本文对均匀光照下以及非均匀光照下图像进行了 实验, 并与一维最小误差法、二维最小误差法、三维直方图重建和降维的Otsu阈值分割 算法、灰度波动变换自适应阈值方法以及一种改进的FCM方法在错误分割率和运行时间上进 行了对比. 实验结果表明, 相对于以上方法, 本算法的分割性能均有明显提升.     

基于复合混沌的鲁棒型医学图像加密算法

当前单一混沌图像加密算法的图像置乱程度低,导致加密结果易受到已知明文攻击,存在安全性低的问题。本文提出基于复合混沌的鲁棒型医学图像加密算法。设置图像密文反馈机制作为算法实现的理论支持,并读取鲁棒型医学原始图像。根据混沌模型的建立条件,构建复合混沌模型。选择Logistics混沌映射,生成复合混沌序列并进行多次迭代后,置乱加密鲁棒型医学初始图像像素值并进行替换与扩散,生成对应的密钥,从而实现鲁棒型医学图像的加密。通过与传统加密算法的对比,本文方法的置乱位数增加了886位,即置乱程度提升了8.7%,且明密文序列相关性较低、抗明文攻击能力较强,实验结果表明该鲁棒型医学图像加密算法的安全性较高。     

基于支持向量机和混沌序列的空域图像双水印

为解决图像同时具有版权保护和内容认证需求问题,提出了一种基于支持向量机的鲁棒水印和混沌序列与LSB相结合的脆弱水印的双重图像水印算法.利用图像邻域像素之间的相关性,通过训练回归型支持向量机模型实现鲁棒水印图像嵌入或提取操作.然后,再将已嵌入鲁棒水印的载体图像用最低有效位和混沌序列相结合的方法嵌入基于载体图像内容的脆弱水印.实验结果表明,该算法同时实现了图像的版权保护和内容篡改定位,提高了水印系统的安全性.     

光照鲁棒的模糊C均值聚类显微图像分割

传统的模糊C均值FCM聚类图像分割算法在显微图像分割中由于没有考虑光照不均匀的影响而降低了分割的效果,为此,提出了一种光照鲁棒的FCM显微图像分割算法。该算法用正交基函数的线性组合模拟不均匀光照,并引入到FCM算法的目标函数中,进行图像的模糊分割。算法不仅降低了非均匀光照对分割效果的影响,还可以同步估计不均匀光照场。实验结果表明,该方法非常有效。     

保持自然度的非均匀光照图像增强的Retinex算法

常规的非均匀照明图像增强方法在增强低光照区域细节时,容易对图像过度增强而导致结果失真.本文从一种新的角度提出了Retinex模型的一种扩展形式,并用于非均匀照明图像的增强.该算法将中心环绕Retinex模型输出作为感知反射率,将图像分解为感知光照图像和感知反射率图像,通过调整感知光照图像,再重新组合感知光照和感知反射率...     

Robust Estimation of Albedo for Illumination-Invariant Matching and Shape Recovery

We present a nonstationary stochastic filtering framework for the task of albedo estimation from a single image. There are several approaches in the literature for albedo estimation, but few include the errors in estimates of surface normals and light source direction to improve the albedo estimate. The proposed approach effectively utilizes the error statistics of surface normals and illumination direction for robust estimation of albedo, for images illuminated by single and multiple light sources. The albedo estimate obtained is subsequently used to generate albedo-free normalized images for recovering the shape of an object. Traditional Shape-from-Shading (SFS) approaches often assume constant/piecewise constant albedo and known light source direction to recover the underlying shape. Using the estimated albedo, the general problem of estimating the shape of an object with varying albedo map and unknown illumination source is reduced to one that can be handled by traditional SFS approaches. Experimental results are provided to show the effectiveness of the approach and its application to illumination-invariant matching and shape recovery. The estimated albedo maps are compared with the ground truth. The maps are used as illumination-invariant signatures for the task of face recognition across illumination variations. The recognition results obtained compare well with the current state-of-the-art approaches. Impressive shape recovery results are obtained using images downloaded from the Web with little control over imaging conditions. The recovered shapes are also used to synthesize novel views under novel illumination conditions.     

三维重建中曲面反照率值的估计与应用

从单幅图像获得物体的表面高度是计算机视觉中的一个重要研究领域,其中一种重要的方法就是从明暗恢复形状(ShapefromShading,简称SFS)。在SFS的各种不同算法中都需要曲面的反照率值,反照率值的估算是否准确直接影响了三维重建的效果。针对反照率值的估算,已经产生了很多有效的算法。文中讨论了三种反照率值的估计算法及其优缺点,并将局部反照率估计算法引入到三维重建中,解决了由全局反照率值重建的弊端。     

一种新的基于线性EIV模型的鲁棒估计算法

提出了一种新的基于线性EIV模型的鲁棒估计算法——鲁棒扩充算法．该算法从结构化数据区域出发,逐渐扩充模型数据集,并不断更新模型参数的估计,直至找到所有模型数据．在每次迭代中,使用C-Step方法对集合进行调整,从而保证了算法的鲁棒性．同时,提出了关于粗差数据和结构化数据分布的结构化密度假设,结合Mean Shift算法,完成对算法的初始位置选取．仿真结果表明,该算法可以有效地处理含有多个结构和大量离群样本的混杂数据,与现有算法相比,具有更强的鲁棒性和更高的精度．     

HMM-Based Gain Modeling for Enhancement of Speech in Noise

Accurate modeling and estimation of speech and noise gains facilitate good performance of speech enhancement methods using data-driven prior models. In this paper, we propose a hidden Markov model (HMM)-based speech enhancement method using explicit gain modeling. Through the introduction of stochastic gain variables, energy variation in both speech and noise is explicitly modeled in a unified framework. The speech gain models the energy variations of the speech phones, typically due to differences in pronunciation and/or different vocalizations of individual speakers. The noise gain helps to improve the tracking of the time-varying energy of nonstationary noise. The expectation-maximization (EM) algorithm is used to perform offline estimation of the time-invariant model parameters. The time-varying model parameters are estimated online using the recursive EM algorithm. The proposed gain modeling techniques are applied to a novel Bayesian speech estimator, and the performance of the proposed enhancement method is evaluated through objective and subjective tests. The experimental results confirm the advantage of explicit gain modeling, particularly for nonstationary noise sources     

A novel algorithm for layout optimization of injection process with random demands and sequence dependent setup times

Injection molding is an ideal manufacturing process for producing high volumes of products from both thermoplastic and thermo setting materials. Nevertheless, in some cases, this type of manufacturing process decelerates the production rate as a bottleneck. Thus, layout optimization plays a crucial role in this type of problem in terms of increasing the efficiency of the production line. In this regard, a novel computer simulation–stochastic data envelopment analysis (CS-SDEA) algorithm is proposed in this paper to deal with a single row job-shop layout problem in an injection molding process. First, the system is modeled with discrete-event-simulation as a powerful tool for analyzing complex stochastic systems. Then, due to lack of information about some operational parameters, theory of uncertainty is imported to the simulation model. Finally, an output-oriented stochastic DEA model is used for ranking the outputs of simulation model. The proposed CS-SDEA algorithm is capable of modeling and optimizing non-linear, stochastic, and uncertain injection process problems. The solution quality is illustrated by an actual case study in a refrigerator manufacturing company.     

Robust filtering of process in the stationary difference stochastic system

A technique to construct the robust Kalman filter for process estimation in the difference linear stationary stochastic system with an unknown covariance observation error matrix was developed. Consideration was given to the algorithm of constructing the set of permissible covariance matrices from a priori statistical data. A numerical method for solution of the general minimax optimization problem was proposed; and on its basis an iterative algorithm to calculate the robust filter parameters was developed, and its convergence was proved. Results of the numerical experiment were presented.     

A Multi-Trip Split-Delivery Vehicle Routing Problem with Time Windows for Inventory Replenishment Under Stochastic Travel Times

This study considers a multi-trip split-delivery vehicle routing problem with soft time windows for daily inventory replenishment under stochastic travel times. Considering uncertainty in travel times for vehicle routing problems is beneficial because more robust schedules can be generated and unanticipated consequences can be reduced when schedules are implemented in reality. However, uncertainties in model parameters have rarely been addressed for the problems in this category mainly due to the high problem complexity. In this study, an innovative and practical approach is proposed to consider stochastic travel times in the planning process. In the planning model, the possible outcomes of vehicle arrivals and product delivery at retailers are systematically categorized and their associated penalty and reward are estimated. Thus, unanticipated costs for every scheduling decision can be incorporated into the planning model to generate vehicle routing schedules that are more robust facing uncertain traffic conditions. To solve the model that is characterized as an NP-hard problem in a reasonable amount of time, a two-stage heuristic solution algorithm is proposed. Finally, the stochastic model is compared with the deterministic model in both planning and simulated operation stages using the data of a supply chain in Taiwan. The result confirms that the schedule generated by the stochastic model is more robust than the one created with the deterministic model because undesired outcomes such as unfulfilled demands are greatly reduced.     

反馈非线性系统随机梯度辨识算法及其收敛性

反馈非线性受控自回归系统是由前向通道的受控自回归模型和反馈通道的静态非线性构成, 这类系统经过参数化后得到双线性参数辨识模型. 本文通过对辨识模型中双线性参数乘积项进行分解, 基于梯度搜索原理, 提 出了反馈非线性系统的随机梯度辨识算法. 为了改善随机梯度算法的收敛速度, 引入遗忘因子, 文章给出了遗忘因子随机梯度算法, 利用随机过程理论, 建立了随机梯度算法的参数估计收敛定理, 证明了算法的收敛性. 最后, 通过数值仿真验证了算法的有效性.     

A neural network controller for systems with unmodeled dynamicswith applications to wastewater treatment

This paper considers the use of neural networks (NN's) in controlling a nonlinear, stochastic system with unknown process equations. The approach here is based on using the output error of the system to train the NN controller without the need to assume or construct a separate model (NN or other type) for the unknown process dynamics. To implement such a direct adaptive control approach, it is required that connection weights in the NN be estimated while the system is being controlled. As a result of the feedback of the unknown process dynamics, however, it is not possible to determine the gradient of the loss function for use in standard (backpropagation-type) weight estimation algorithms. In principle, stochastic approximation algorithms in the standard (Kiefer-Wolfowitz) finite-difference form can be used for this weight estimation since they are based on gradient approximations from available system output errors. However, these algorithms will generally require a prohibitive number of observed system outputs. Therefore, this paper considers the use of a new stochastic approximation algorithm for this weight estimation, which is based on a "simultaneous perturbation" gradient approximation. It is shown that this algorithm can greatly enhance the efficiency over more standard stochastic approximation algorithms based on finite-difference gradient approximations. The approach is illustrated on a simulated wastewater treatment system with stochastic effects and nonstationary dynamics.     

Functional series modeling approach to identification ofnonstationary stochastic systems

The problem of identification of a nonstationary stochastic system is considered, and an estimation method based on the functional series modeling (FSM) of the system parameter trajectory is proposed for its solution. It is shown that the parameter-matching properties of FSM estimators can be described in terms of the appropriately defined (time-varying) impulse and frequency responses. It is suggested and verified by means of computer simulation that the averaged frequency characteristics associated with FSM estimators can yield useful information about their parameter-matching abilities