自适应尺度的双边纹理滤波方法

目前已有的结构保持的纹理平滑方法主要是利用矩形片内的统计量来区分纹理和结构,但是所用的矩形片边长是单一尺度的,这将导致含有尖锐结构或结构在多个尺度上的图像出现纹理过平滑或未平滑的现象。为此,提出一种自适应尺度的双边纹理滤波方法。首先,通过对局部区域进行统计分析,从给定候选值中自适应地为每个像素选取合适的矩形片边长,对于均匀的纹理区域,选取较大的矩形片边长,对于邻近特征边的区域选取较小边长;其次,利用自适应的矩形片边长计算引导图像;最后,对原始图像进行引导双边滤波。实验结果表明,所提方法能够在保持图像结构的同时更好地平滑纹理。     

一种自适应图像去噪混合滤波方法

结合自适应中值滤波技术和自适应压缩加权均值滤波技术,提出了一种新的图像混合噪声滤波算法。算法首先对受混合噪声污染的图像利用灰度极值检测出脉冲噪声,运用自适应中值滤波滤除脉冲噪声;其次对处理结果进行自适应压缩的加权均值滤波。实验结果说明算法不仅能有效地滤除脉冲与高斯混合噪声,而且可以较好地保护图像细节。     

雪深超声回波信号的自适应滤波建模方法

针对现阶段获取积雪数据准确度不高的问题,根据超声回波信号特征并结合建模方法和现代数字信号处理技术,提出一种积雪超声回波信号的自适应滤波建模方法,利用包络检波算法和曲线拟合对回波信号包络中的确定性信号进行拟合并结合自适应滤波算法,在误差信号的反馈下通过迭代使输入信号逼近参考信号,改变输入信号幅值和频率也可以稳定地收敛。通过这种方法可以合成积雪回波信号,准确描述雪盖的层理结构,有效地降低了噪声干扰,拟合匹配度得到提高。     

图像椒盐噪声的自适应滤波算法研究

为有效去除严重的椒盐噪声、更好地保护图像细节,提出了一种基于改进脉冲耦合神经网络（PCNN）的自适应去噪方法。根据PCNN神经网络的点火时刻矩阵,对受噪声污染的像素进行定位,仅对噪声像素进行类中值滤波,实现了图像细节的有效保留;根据噪声强度的估计信息,自动进行滤波次数和滤波窗口尺寸的优选,实现了图像的强自适应滤波。实验表明,与传统去噪方法相比,该方法噪声去除效果好,图像细节保持完整,而且系统具有一定的泛化能力。     

基于脉冲耦合神经网络的自适应图像滤波

为有效滤除灰度图像中的椒盐噪声并保留图像的边缘及细节信息,提出一种简化的阈值单向衰减脉冲耦合神经网络(PCNN)点火矩阵自适应图像滤波方法,简化的PCNN结构减少了所需参数并提高了运算速度。该方法通过对PCNN点火矩阵的分析,定位出被噪声污染的像素,只对噪声像素进行滤波,因而有效地保留了图像的细节信息;并根据椒盐噪声的特点,动态估计图像的噪声强度,自适应地选择滤波窗口的大小和滤波次数。实验结果表明提出方法较常见的图像降噪方法在滤波效果、自适应性及保留图像细节方面有明显的优势。     

一种结合小波阈值滤波的自适应整体变分方法

在整体变分方法去噪原理的基础上,通过引入小波阈值滤波,用自适应正则项代替整体变分模型中的正则项,提出了一种依赖于信号的局部信息进行滤波的自适应整体变分方法,自适应地在整体变分正则化和各向同性光滑化之间调整滤波强度。为求解整体变分极小化问题,采用了滞后扩散定点迭代的方法。数值计算结果表明：提出的方法有效地减少了传统整体变分方法去噪后恢复信号中所出现的阶梯效应,很好地抑制了小波变换中固有的伪Gibbs现象,重构信号的边缘、不连续点位置十分精确,信噪比也得到明显改善。     

Adaptive anti-spam filtering for agglutinative languages: a special case for Turkish

We propose anti-spam filtering methods for agglutinative languages in general and for Turkish in particular. The methods are dynamic and are based on Artificial Neural Networks (ANN) and Bayesian Networks. The developed algorithms are user-specific and adapt themselves with the characteristics of the incoming e-mails. The algorithms have two main components. The first one deals with the morphology of the words and the second one classifies the e-mails by using the roots of the words extracted by the morphological analysis. Two ANN structures, single layer perceptron and multi-layer perceptron, are considered and the inputs to the networks are determined using binary model and probabilistic model. Similarly, for Bayesian classification, three different approaches are employed: binary model, probabilistic model, and advanced probabilistic model. In the experiments, a total of 750 e-mails (410 spam and 340 normal) were used and a success rate of about 90% was achieved.     

Adaptive smoothing: a general tool for early vision

A method to smooth a signal while preserving discontinuities is presented. This is achieved by repeatedly convolving the signal with a very small averaging mask weighted by a measure of the signal continuity at each point. Edge detection can be performed after a few iterations, and features extracted from the smoothed signal are correctly localized (hence, no tracking is needed). This last property allows the derivation of a scale-space representation of a signal using the adaptive smoothing parameter k as the scale dimension. The relation of this process to anisotropic diffusion is shown. A scheme to preserve higher-order discontinuities and results on range images is proposed. Different implementations of adaptive smoothing are presented, first on a serial machine, for which a multigrid algorithm is proposed to speed up the smoothing effect, then on a single instruction multiple data (SIMD) parallel machine such as the Connection Machine. Various applications of adaptive smoothing such as edge detection, range image feature extraction, corner detection, and stereo matching are discussed     

解大规模无约束优化的自适应过滤信赖域法

提出一种解大规模无约束优化问题的自适应过滤信赖域法。用目标函数的梯度及迭代点的信息来构造目标函数海赛矩阵的近似数量矩阵,引进了过滤技术和自适应技术,大大提高了计算效率。从理论上证明了新算法的全局收敛性,数值试验结果也表明了新算法的有效性。     

自适应模型更新相关滤波目标跟踪方法

为了解决目标因遮挡、跟踪框发生漂移后相关滤波跟踪算法仍持续更新目标模型和滤波器模型,导致背景信息被更新到目标模型和滤波器模型中的情况,提出一种自适应模型更新策略。采用相关滤波方法得到新的目标位置;提取新目标位置的统计协方差特征并计算其与协方差模板的相似性;根据相似性判断是否更新目标模型和滤波器模型。实验结果表明:所提方法有效解决了因目标形变、遮挡等情况导致目标模型和滤波器模型的更新问题,提高了相关滤波目标跟踪的精度。     

Adaptive nonlocal filtering: a fast alternative to anisotropicdiffusion for image enhancement

Nonlinear anisotropic diffusion algorithms provide significant improvement in image enhancement as compared to linear filters. However, the excessive computational cost of solving nonlinear PDEs precludes their use in real-time vision applications. We show that two orders of magnitude speed improvement is provided by a new image filtering paradigm in which an adaptively determined vector field specifies nonlocal application points for an image filter     

A general method for the shape finding of lightweight tension structures

A computer based procedure for finding the shape of tension roof structures is presented. The method is quicker, more accurate and cheaper than that using models. A number of examples taken from existing and proposed large scale structures are given.     

Filter shape dependence and effective scale separation in large-eddy simulations based on relaxation filtering

The influence of the filter shape on the effective scale separation and the numerical accuracy of large-eddy simulations based on relaxation filtering (LES-RF) is investigated. The simulation of the turbulent flow development of a high-Reynolds number low-subsonic compressible mixing layer is performed using the LES-RF procedure, for discrete filters of order 2–10. A reference solution is first obtained using high-order numerical algorithms and shows a good agreement with experimental data found in the literature. Discrete filters of order 2, 4, 6, 8 and 10 are then considered to study the influence of the filter shape on numerical results. The 2nd-order scheme turns out to be too dissipative and prevents the emergence of unsteady motions within the mixing layer. For higher order schemes, from 4th- to 10th-order, the flow solutions are turbulent but exhibit mean flows and turbulent intensities depending on the filter. The investigation of the one-dimensional kinetic energy spectra then shows that the 4th-order filter may still be too dissipative whereas large scales remain unaffected using the 6th-, 8th- and 10th-order filters. A further study of the kinetic energy spectra nonetheless demonstrates that the effective spatial bandwidth of the LES increases with the order of the filtering scheme. Simulations using the 6th-, 8th- and 10th-order filters, with mesh sizes chosen to provide the same effective LES cut-off wavenumber, are performed and yield similar results. It is hence found that the value of the effective LES cut-off wavenumber, rather than to the filter shape itself, is mainly responsible for the discrepancies between the flow statistics obtained using different filters. One may conclude that filter shape independence is consequently achieved in the present LES of a mixing layer.     

基于快速判别式尺度估计的核相关滤波方法

针对核相关滤波（KCF）算法无法对视频序列中目标尺度变化作出响应的问题，提出一种基于快速判别式多尺度估计的核相关滤波跟踪算法。首先，使用核相关滤波器来估计目标位置；然后，通过使用一组不同尺度的目标样本来在线学习快速判别式尺度滤波器；最后，在目标位置应用学习的尺度滤波器来获得目标尺寸的准确估计。选取Visual Tracker Benchmark视频序列集进行实验，并与基于判别式尺度空间跟踪（DSST）的KCF算法和传统KCF算法进行对比，结果表明，在目标尺度发生变化时，所提算法在跟踪精度上提高了2.2%至10.8%；并且在平均帧率上，所提算法比DSST的KCF算法提高了19.1%至68.5%，表明该算法对目标尺度变化有很强的适应能力和较高的实时性。     

动态贝叶斯网络一种自适应的局部抽样粒子滤波算法*

针对传统自适应粒子滤波（APF）对于动态贝叶斯网络推理中高维的问题,提出动态贝叶斯网络一种自适应的局部抽样粒子滤波算法(LSAPF)。LSAPF算法将BK算法分团的思想引入到粒子抽样中,利用策略相关性和局部模型的弱交互性为指导对动态贝叶斯网络进行分割,以降低抽样规模和抽样的状态空间;进而对局部模型用自适应粒子滤波算法进行近似推理,并以粒子的因式积形式近似系统的状态信度。实验结果表明,该算法能很好地兼顾推理精度和推理时间,其性能优于普通PF算法;与APF算法相比,在不增加推理误差的情况下推理时间也有较大的提高。     

A spatio-frequency trade-off scale for scale-space filtering

We study implementation issues for spatial convolution filters and their Fourier alternative, with the aim to optimize the accuracy of filter output. We focus on Gaussian scale-space filters and show that there exists a trade-off scale that subdivides the available scale range into two subintervals of equal length. Below this trade-off scale Fourier filtering yields more accurate results than spatial filtering; above it is the other way around. This should be contrasted with demands of computational speed, which show the opposite tenet     

An efficient method for obtaining the general solution for theforce balance equations with hard point contacts

A compact formulation which is more efficient than the traditional pseudoinverse formulation for obtaining the general solution for the force balance equations has been presented (Cheng and Orin, 1991). With hard point contacts considered, the force balance equations can be decomposed into two sets of rank 3, smaller linear equations if proper coordinate frames of the reference member and at the contact points are chosen. This decomposition, together with the compact formulation, can reduce the steps of the Gaussian elimination process, increase parallelism of the algorithm, and therefore, keep the computation time for obtaining the general solution for the force balance equations to a minimum     

一种盲信号处理中的时域自适应滤波算法

滤波是信号处理中的重要环节,鉴于盲信号处理本身的特点,传统的滤波技术并不适合直接用于盲源分离之中。然而作为分离前的预处理,滤波技术在独立成分分离算法中是必要的。为此,本文结合稳健的数据非线性投影,首次提出盲信号中的自适应滤波方法,与此同时给出了具有自适应特性的阈值判决。在此基础上构造了盲信号中的自适应滤波算法,解决了利用低通和高通滤波处理盲信号所遇到的问题。仿真结果表明,在不破坏数据统计特性的前提下,该方法能有效滤除数据中的野值成分,避免了野值数据对独立成分分离算法性能的影响,为盲信号分离的预处理开辟了一种新的途径。     

Adaptive Gaussian weighted filtering for image segmentation

A simple approach to image smoothing is presented using a filter whose weights are adaptive Gaussian functions of the absolute difference between the central pixel and its neighbours in a 3 × 3 window. Such a filter smooths the signal effectively within homogeneous regions whilst maintaining inter-region boundaries thus alloqing iterative processing. The filtering scheme is shown to be particularly suitable for image segmentation.