简析网络不良内容的检测和屏蔽手段

网络不良内容的辨析和屏蔽已越来越引起人们地重视。本文通过对图片内不良内容辨析与过滤、不良视频检索两个关键技术领域进展情况的梳理,提出综合运用分割时域、提取关键帧、识别图像内容及检测皮肤等技术问题的解决途径,并且阐述了当前网际不良内容识别和屏蔽的研究情况和主要应用,分析了其存在的问题。     

导波雷达液位计的回波信号处理电路设计

该文简要地介绍了雷达液位计的分类和导波雷达液位计的测量原理,在此基础上设计出导波雷达液位计的回波信号处理电路,并详细地分析了回波信号的处理过程。所设计的电路简单实用,而且系统运行稳定,能满足工业测量要求。     

数据库汉语查询语言的分词研究与实现

在综合考虑数据库查询这一特殊性的基础上,根据查询语句中词汇对数据查询不同贡献程度分级建立分词词典;然后提出了分步- - 正向单扫描的分词方法(DSWS) ,并对该分词方法的时间复杂度进行了分析。     

视频序列图像中运动对象分割综述*

视频分割技术作为多媒体技术应用的一个主要方面, 在视频编码、检索、多媒体交互中存在着广阔的前景, 而且在计算机视觉中也有着极为重要的应用; 但是由于视频图像内容复杂, 还没有一个通用的分割方法, 视频分割算法还面临着挑战。主要介绍视频分割中采用的一些算法, 并对各种算法和性能进行了比较和讨论。     

基于空间约束Student’s-T混合模型的模糊聚类图像分割

针对基于高斯混合模型的模糊聚类算法对噪声和异常值敏感的问题,利用包含邻域关系的先验概率与Student’s-T分布构建基于空间约束的混合模型,并结合熵规则化项定义模糊聚类目标函数。Student’s-T分布具有重尾的特点,较之高斯分布具有更强的抗噪能力。此外,为了更加有效地平滑噪声,在标号场上利用马尔科夫随机场模型刻画包含像素与其邻域像素相关性的先验概率,并表达为混合模型的权值系数以增强算法的鲁棒性。通过对模拟图像和真实彩色图像分割结果的定性定量分析,验证了提出算法的有效性和可行性。     

一种对光照具有鲁棒性的图像分割方法

研究了灰度图像的OTSU(最大类间方差)自动阈值分割法。OTSU方法作为一种单一阈值的分割方法,当图像受光照和反射光等的影响明显时,将会出现严重的误分割现象。考虑到OTSU方法的最大类间方差化的思想,根据灰度图像的像素点灰度的直方图分布、空间分布,提出了一个新的分割阈值方法。先根据OTSU方法的特点自设计一个函数,对图像进行变换,以便后面的处理,再对其图像以改进的OTSU方法进行分割。通过对化学实验中两种液体的拍摄图片及数字图像处理中标准图片进行试验,理论分析与实验结果表明：该方法能够对受光照及反射光影响大的图像实现正确的分割,将目标图像清晰地从背景中分割出来。     

基于精度依据准则的图像分割算法评价

在图像处理中,图像分割是一类重要的研究方向。图像分割算法的好坏,影响到分割结果的优劣,因此对分割算法的性能评估十分重要。本文提出了一种图像分割算法性能的评价方法——精度依据准则,该准则是对原始特征量值和实际特征量值做比较,通过对绝对值的大小来判断算法的好坏。通过实验比较,该方法具有不错的算法性能评估准确度。     

中文分词技术在搜索引擎中的应用

本文介绍了搜索引擎的关键技术——中文分词技术,对中文分词技术的概念、常用算法以及分词的难点进行了详细介绍,让读者对中文分词的现状有所了解。     

融合深度和颜色信息的图像物体分割算法

讨论立体图对的图像分割问题,提出一种基于深度和颜色信息的图像物体分割算法。该算法首先利用基于聚类的Mean-shift分割算法对目标图像进行适度的过分割,同时借助双目立体视觉算法获取立体图对的稠密深度图,并依据深度不连续性从过分割结果中选取用于继续进行“精致”分割的种子点集,接着对未分配种子标签的区域用图割算法分配标签,并对彼此之间没有深度不连续边界但具有不同标签的相邻区域进行融合。相比于传统图像分割算法,该算法可有效克服过分割和欠分割问题,获取具有一定语义的图像分割结果。相关的对比实验结果验证了该算法的有效性。     

Time-domain decomposition of optimal control problems for the wave equation

We consider the problem of boundary optimal control of a wave equation with boundary dissipation by the way of time-domain decomposition of the corresponding optimality system. We develop an iterative algorithm which shows that the decomposed optimality system corresponds to local-in-time optimal control problems which can be treated in parallel. We show convergence of the algorithm. Finally, we provide a time discretization which is reminiscent of an instantaneous control scheme. We thereby also contribute to the problem of convergence of such schemes.     

Measurement of three independent components in impedance sensors using a single square wave

Two-wire impedance-based sensors involving electrolytes add the impedance of the electrodes to the electrical impedance of the medium to measure. An equivalent circuit for the measured impedance is a resistance in series with the parallel combination of another resistance and a capacitance. If the two electrodes are modelled by equal impedances, the equivalent circuit for the complete set up consists of three impedance components, which can be determined from three independent measurements. This paper describes a novel method to obtain those three components using a single square wave voltage (period 2T) instead of several sine waves and provides the equations to calculate their value from the three current intensity amplitudes measured at T/8, 3T/8 and 5T/8. Other measurement times would need different equations, but the same procedure applies. Anyway, the proposed method keeps the advantages of synchronous detection and relies on analytical solutions instead of the customary curve fitting procedures. Computer simulation and experimental results obtained by measuring the conductivity of known electrolyte samples validate the proposed method.     

一种基于数据库的图像分割评价方法*

介绍了一个专用于机票工程项目中分割算法评测方法。该方法基于两个图像数据库及一套误差度量准则。图像数据库中的标准图像部分是基于主观视觉一致性的人工分割,具有评价的统计客观性。初步的应用证明该研究尤其是对算法的细微分析、提高整个工程有效性具有重要意义。     

基于最短路径的二元语法中文词语粗分模型的研究

词语粗分是分词后续处理的基础和前提,直接影响到分词系统最终的准确率和召回率。针对目前常用分词方法单一使用时存在的不足,综合机械分词的高效性和统计分词的灵活性,设计一种基于最短路径的二元语法中文词语粗分模型。实验结果表明,此粗分模型无论在封闭测试和开放测试中,还是在不同粗分模型对比测试和不同领域的开放测试中,都有较好的句子召回率。     

A microcontroller system for measurement of three independent components in impedance sensors using a single square pulse

A novel time domain method and its implementation in a simple smart impedance sensor controlled by an 8-bit microcontroller is presented in the paper. The method is based on stimulation of a voltage divider consisting of a resistor working as a current-to-voltage converter and the impedance sensor by a single square-voltage pulse with a duration time T directly generated by the output of the microcontroller and on sampling the resulting voltage on the sensor at three different selected moments T/8, T/2 and 7T/8 by the internal ADC of the microcontroller. The sensor is modeled by a three-components circuit. The duration time T is determined by the first timer and the ADC is triggered by the second timer of the microcontroller. The measurement procedure takes less than 1 ms and the determination of model component values is based on basic calculus. Thanks to this, smart sensors basing on this solution are energy-saving, they can work on the same batteries by a few years, and low cost on the level tens euros. Hence, they can be used in wireless sensor networks, especially basing on the ZigBee protocol. The results of the simulation investigation and the experimental verification of the method are included in this paper.     

图像分割研究进展

至今提出的分割算法有上千种,而新的算法还在不断被提出。从图像分割实际应用的角度,或者从图像分割中使用的特定的理论工具的角度,可以对图像分割进行分类。Otsu算法、基于聚类的图像分割方法、运动分割、基于图论的图像分割方法和基于活动轮廓的图像分割方法是图像分割的主要研究领域,自然计算算法已成为图像分割新的研究热点。     

Albedo estimation for scene segmentation

Standard methods of image segmentation do not take into account the three-dimensional nature of the underlying scene. For example, histogram-based segmentation tacitly assumes that the image intensity is piecewise constant, and this is not true when the scene contains curved surfaces. This paper introduces a method of taking 3D information into account in the segmentation process. The image intensities are adjusted to compensate for the effects of estimated surface orientation; the adjusted intensities can be regarded as reflectivity estimates. When histogram-based segmentation is applied to these new values, the image is segmented into parts corresponding to surfaces of contant reflectivity in the scene.     

一种粘连颗粒图像的快速分割方法

对于工业和医学中常见的颗粒图像，提出了一种快速分割粘连颗粒的方法。利用颗粒图像的轮廓特征，找到若干拐点，从拐点中进一步提取出分割点。然后利用分割点的相互距离等因素，为每个分割点找到相匹配的分割点。连接每对分割点即可分开相粘连的颗粒。这种方法快速有效，取得了良好的效果。     

基于前向平均和时域滤波的OFDM信道估计算法

为减少系统中噪声对信道估计的影响,提出了一种基于抑制噪声的信道估计方法.首先对几个连续0FDM信号相同导频位置的信道频率响应前向平均后线性插值,得到所有子载波的信道估计值;然后在时域利用滑行相关系数估计出的时延进行时域滤波.该方法通过在频域和时域的两次减噪,提高了信道估计的精确性,尤其是所提出的前向平均法适用于快衰落信道.仿真结果表明,该方法可将OFDM系统的误码率性能提高4～5 dB.     

汉语自动分词研究及其在信息检索中的应用

汉语自动分词问题是中文信息检索的基础问题，也是阻碍其向前发展的“瓶颈”问题。介绍了分词技术的发展状况，分析了分词技术在信息检索过程中的应用，并讨论了信息检索和分词技术结合的方式和需要以及解决的关键技术问题。     

Segmentation of ARX-models using sum-of-norms regularization

Segmentation of time-varying systems and signals into models whose parameters are piecewise constant in time is an important and well studied problem. Here it is formulated as a least-squares problem with sum-of-norms regularization over the state parameter jumps, a generalization of ?₁-regularization. A nice property of the suggested formulation is that it only has one tuning parameter, the regularization constant which is used to trade-off fit and the number of segments.