基于重构样本区域的数字图像修复算法研究

基于样本块的纹理合成修复算法（Criminisi算法）在匹配样本块时,因为使用全局搜索策略而降低了有效匹配效率。针对这一问题,重点研究了一种改进的基于样本块的数字图像修复算法。改进算法考虑到相邻像素块的相似性,对图像样本区域进行了重构,减少了样本块搜索范围,提高了图像修复效率。     

Joint time-frequency synchronization and channel estimation in two-way relay networks

This work investigates joint time-frequency synchronization and channel estimation in two-way relay networks (TWRN) that utilize amplify-and-forward (AF) relay strategy. A two-way transmission model for joint time-frequency synchronization and channel estimation in TWRN is provided. Due to the high dimension and complexity of the model, the maximum likelihood (ML) algorithm is over complex and a standard Markov chain Monte Carlo (MCMC) algorithm is ineffective. We instead develop a hybrid Metropolis–Hastings–Gibbs algorithm in order to facilitate joint time-frequency recovery and effective channel estimation. In particular, we present a reparameterized model to facilitate Gibbs sampling with respect to the self-interference channel. Then the second order truncated Taylor series approximation is adopted for carrier frequency offsets (CFOs) estimation. Furthermore, a heuristic way of determining update order of parameters is proposed and initial value of the Markov chain is discussed as well. To test the robustness and objectivity of our proposed algorithm, two MCMC estimators, each with its own prior distribution, are compared against each other. Numerical results are provided to verify the effectiveness and robustness of proposed algorithm in terms of both mean-square errors (MSE) and estimation bias. The bit error rate (BER) performance results are further offered for comparison.     

改进的K-均值聚类算法图像边缘检测研究

图像边缘检测一直是图像处理领域研究的重点问题。边缘是图像最基本的特征,本文采用了模糊K-均值聚类算法对图像进行边缘检测。该方法针对不同的图像找到相对比较有效的边缘检测算法,进而大幅度地减少了数据量,保留了图像重要的结构属性。通过mat lab实验,证明了该方法可以有效提取图像的边缘信息。     

基于爬山法快速波形匹配的高频电源过程控制

提出了基于波形匹配的方法预测一次电流过零点,来提高高频电源输出功率.然而直接使用归一化互相关算法计算量大,速度慢.为了提高匹配速度,本文采用爬山法启发式搜索策略,实现了非遍历性搜索,并能快速找到最优匹配点.实验结果表明,爬山搜索算法性能优越,用于高频电源过程控制中是可行的,且具有快速性和较高的准确性,相比直接互相关算法...     

基于SIFT特征点改进搜索算法的图像匹配技术研究

本文研究一种改进的近邻搜索算法的图像匹配技术。本文采用基于特征的图像匹配方法,利用SIFT算法提取特征点。在特征点匹配的过程中,为提高搜索样本特征点的最近邻和次近邻特征点的速度,本文采用一种基于二叉检索树算法改进的近邻搜索算法,该算法用最近邻与次近邻比值来进行特征点的匹配。用MATLAB语言实现该算法并运用到图像特征匹配中,实验证明优于原算法并具有较高实时性。     

基于中值背景模型的运动目标自适应检测方法

针对复杂交通场景中运动车辆检测方法存在的局限性,本文提出了一种基于中值模型和自适应阈值的运动检测算法。利用自适应阈值对差分图像的三个颜色通道进行二值化处理,实现了运动目标的精确检测,采用中值更新策略实现背景图像的实时更新。实验结果表明,算法可以从复杂交通场景图像序列中有效地检测出运动目标,且算法计算量小,具有良好的鲁棒性与实时性。     

A stochastic multi-parameters divergence method for online auto-tuning of fractional order PID controllers

This paper presents a stochastic multi-parameters divergence method for online parameter optimization of fractional-order proportional–integral–derivative (PID) controllers. The method is used for auto-tuning without the need for exact mathematical plant model and it is applicable to diverse plant transfer functions. The proposed controller tuning algorithm is capable of adaptively responding to parameter fluctuations and model uncertainties in real systems. Adaptation skill enhances controller performance for real-time applications. Simulations and experimental observations are carried on a prototype helicopter model to confirm the performance improvements obtained by the online auto-tuning of fractional-order PID structure in laboratory conditions.     

基于Adaboost的人脸光照补偿方法

针对传统Adaboost人脸检测算法中光照不均匀、图像检测正确率低的问题,提出一种基于Adaboost算法的人脸光照补偿检测方法。介绍Adaboost人脸检测算法的处理流程以及直方图均衡化原理,并将Adaboost算法和直方图均衡化相结合,实现人脸检测。检测结果表明,与传统的Adaboost人脸检测方法相比,新方法对于光照不均匀图像的人脸检测有很好的效果。     

基于改进的LBP特征和随机森林相结合的人脸关键点检测方法研究

提出了一种人脸关键点检测方法,该方法用了少量的正面图像,不用归一化人脸图像,而传统的人脸关键点检测方法需要对图像进行严格预处理。随机森林是一种分类器融合算法,可以很好地解决多类分类问题,虽然LBP特征简单,但其可以包含大量的纹理信息。利用改进的LBP特征与随机森林相结合,构成一种对人脸关键点检测的方法。通过高斯平滑图像的LBP特征的提取,对每个点生成特征,计算出有用的特征作为正例,并且与反例集合变为训练集。通过随机森林分类器进行分类,误差率较低,仅在10％左右。     

针对纹理图像压缩的改进SPIHT算法

针对纹理图像压缩提出了一种改进的SPIHT算法. 算法对一阶小波变换产生的LL、HL、LH、HH 4个子带系数分别进行小波再分解,之后对上述4个子带利用SPIHT算法进行联合编码;子带之间的量化比特分配嵌入在编码过程之中,因此可实现4个子带编码比特的自动精确分配. 编码过程充分利用了小波系数的高频分量信息,可有效保持图像中的纹理细节. 最后通过对纹理图像的压缩实验,证实了改进算法的有效性.     

基于模糊C-均值聚类图像分割技术研究

主要研究了均值聚类图像分割问题。针对传统的聚类图像分割算法对图像地分割精度较低等问题,提出一种基于模糊控制的C-均值聚类快速图像分割新方法。本文采用快速模糊C-均值聚类算法对图像分割。实验结果表明,图像分割边缘清晰,分割效果明显优于传统的聚类图像分割算法。     

一种改进的基于马尔科夫链的扩频图像隐写分析方法

提出一种对基于扩频信息隐藏图像进行隐写分析的方法.应用马尔科夫链模型,根据图像的相邻像素点之间的相关性,结合阈值判别方法和改进的模式识别方法,判断一幅图像中是否存在隐写信息.在Corel图像库上的实验表明,本方法降低了虚警率,且正检率的性能优于已有方法.     

A robust super-resolution method with improved high-frequency components estimation and aliasing correction capabilities

In this paper, we have proposed a robust super-resolution high-frequency component estimation (RS-HFCE) method, which can efficiently estimate lost high-frequency components and correct aliasing effects of low-frequency components of an image. The fundamental principle of operation of the proposed method is based on the idea that, when a baseband band-limited image signal of known bandwidth in a high-resolution lattice is iteratively low-pass filtered in the frequency domain, the unknown values in the lattice can be interpolated, thus correcting the aliasing for the low-frequency components. If this process is done along with adjusting the amplitudes of the known pixel values, some high-frequency components of an image are automatically extrapolated. In order to provide simultaneous edge preservation and noise removal capabilities of the super-resolved images, an improved version of an adaptive Perona–Malik (PM) model is incorporated into the process. One of the characteristics of the proposed method is its high level of tolerance capabilities to reconstruction errors and noise caused by an increase in the reconstruction scaling factors. High quality images of higher resolution are still appreciably reconstructed when greater magnification factors are used. From a couple of experiments on real images, and using both subjective and objective image quality assessment measures, it is demonstrated that the proposed method outperforms most of other classical methods.     

基于混沌和分维的盲水印算法

把分形维数理论应用到数字水印中,提出了基于Chebysher混沌置乱和分形维数的自适应数字图像水印算法。首先将载体图像分块,计算每个小块的分形维数。然后将载体图像进行分块DCT变换,使用改进的邻域平均法,将经过混沌置乱后的水印信息嵌入到图像的DCT域中,并根据该小块的分形维数调节嵌入强度,实现了水印信息的自适应嵌入。提取水印时,实现了完全盲提取。MATLAB仿真结果表明,该算法具有较好的不可见性,对常见的图像处理攻击具有较好的鲁棒性。     

基于混合自适应的遗传工作流挖掘算法

针对遗传算法工作流挖掘容易过早收敛且局部寻优能力较差,导致得到的解不理想的情况,提出了一种基于混合遗传方法的工作流挖掘算法。该算法采用因果矩阵映射流程实例作为工作流模型的编码,在遗传算法的选择操作阶段采用锦标赛策略与精英保留策略相结合,在交叉变异阶段运用混合自适应方法,并结合模拟退火思想,使解的质量有了明显的提高。仿真实验表明,该算法与基于简单遗传方法的工作流挖掘算法相比效率更高。     

基于改进遗传算法的高光谱图像波段选择

在对地观测领域,高光谱图像得到了广泛应用,但存在数据量大、波段间相关性高等问题. 针对以上问题分析了已有的波段选择方法,提出了基于信息量及类间可分离性准则的遗传算法对高光谱图像进行波段选择:构造波段互相关系数矩阵进行子空间划分;利用联合熵作为组合信息量的标准,Bhattacharyya距离作为类间可分离性标准,构造遗传算法的适应度方程,改进了遗传算法中的选择算子. 最后用AVIRIS图像对提出的算法进行试验,并利用最大似然分类法对最优波段组合进行分类,总体分类精度达到94.24%,Kappa系数达到0.94.     

智能小车中图像采集和处理的研究和实现

以智能小车为载体,研究了在实时性需求较高的环境中图像采集和图像处理无法交互工作而带来的低效率问题,提出了一种利用图像采集间隙将图像识别同步到图像采集中以提高效率的方法;针对实时系统中图像采集时的噪声,反光等各种干扰,本文还讨论了赛道图像的滤波去噪方法,同时在固定阈值提取赛道中心线方法的基础上,研究了一种动态阈值法提取赛道中心线,进而给出可靠性高的赛道信息供给控制单元做下层调理。实验表明,该方法能够很好地应用于智能小车等实时性较高的图像采集和图像处理系统中。     

一种彩色图像的边缘检测算法

通过对灰度图像的边缘检测的研究,构造了一种新的基于色度差的边缘检测算法,并充分利用彩色图像的颜色信息,将此算法从灰度图像转化到RGB的颜色空间中。这种新方法旨在区别于传统意义上对图像的边缘检测要求的精准性,而把提取出彩色图像中直观形象的轮廓信息作为研究目的。实验仿真表明,该算法提取出的边缘能够较好地反映目标图像中具有代表性的信息。     

Improved sentence retrieval using local context and sentence length

In this paper we propose improved variants of the sentence retrieval method TF–ISF (a TF–IDF or Term Frequency–Inverse Document Frequency variant for sentence retrieval). The improvement is achieved by using context consisting of neighboring sentences and at the same time promoting the retrieval of longer sentences. We thoroughly compare new modified TF–ISF methods to the TF–ISF baseline, to an earlier attempt to include context into TF–ISF named tfmix and to a language modeling based method that uses context and promoting retrieval of long sentences named 3MMPDS. Experimental results show that the TF–ISF method can be improved using local context. Results also show that the TF–ISF method can be improved by promoting the retrieval of longer sentences. Finally we show that the best results are achieved when combining both modifications. All new methods (TF–ISF variants) also show statistically significant better results than the other tested methods.