Facial detection and location based on skin color and eye characteristic

This paper comes true a system of facial detection and location. Firstly, it realizes the facial detection of candidate region used skin color for the given image, and considers the influence of similar skin. Secondly, it realizes the location for eye region used the gray characteristic of eye and correction of central measurement, and makes further selection for facial candidate region. The end, it normalizes the facial image and makes it as front standard image, so realizes the facial characteristic location.     

低压补偿滤波装置的应用分析

用实例对比了低压补偿滤波装置与传统的无功补偿装置对工业用户无功补偿的作用和安装，运行、维护情况，阐明了低压补偿滤波装置是一种兼顾了补偿和滤波功能，具有较高的技术性和经济性的无功补偿和谐波滤波装置。     

机载红外搜索跟踪系统被动定位滤波算法研究

首先用扩展卡尔曼滤波算法构建了机载红外搜索跟踪系统被动定位滤波模型．然后针对该滤波算法要求先验的噪声统计及存在系统观测模型线性化误差影响滤波精度的特点．利用虚拟噪声技术，提出了适合于红外搜索跟踪系统被动定位的自适应扩展卡尔曼滤波算法。该算法实时地估计了虚拟噪声的统计特性，减小了线性化误差，提高了非线性滤波的精度。仿真结果表明，在完全相同的初始条件下，自适应扩展卡尔曼滤波对目标距离和速度的估计结果明显优于扩展卡尔曼滤波，此算法具有很高的工程应用价值。     

二维卡尔曼─中值混合自适应滤波

本文提出了能同时抑制加性噪声，脉冲噪声干扰和点扩展函数模糊，又能保护图象边缘及细节信息的新型滤波方法—“二维Kalman－Median混合自适应滤波”，并成功地应用于统计图象恢复，取得了满意的效果。     

Influence of zero-loss filtering on electron optical phase contrast

Complex scattering amplitudes are used to calculate the phase contrast of colloidal gold particles. Comparison of measurements of the phase contrast intensity at the centre of the gold particle as a function of defocus for unfiltered and zero-loss filtered images demonstrates the increase in phase contrast achieved by zero-loss filtering even for a thick carbon substrate film. The granulation of amorphous germanium films is measured by the spatial rms (root mean square) values of image intensity in a defocus series.     

实时跟踪控制系统中的数字滤波技术

在高精度的实时跟踪控制系统中，由探测器等其他一些因素给系统带来的噪声将不容忽视，数字滤波将是滤除噪声的有效手段，，由于系统实时性和采样频率的要求，本文介绍几种时跟踪系统中常用的数字滤波方法，并对它们在实时跟踪系统中的作用进行了和比较。     

DNS／GPS组合导航系统研究

本文阐述ＤＮＳ与ＧＰＳ组合导航的必要性、可行性及组合方法，给出了误差模型和卡尔曼滤波模型。选取观测量不同可有三种组合方式，通过模拟计算给出了相应曲线。当采用伪距及其变化率作为观测量时，其组合方式最佳，定位精度可达米级，航向精度可提高到０．２°。     

一种改进的网格数据保持梯度滤波方法

 基于网格数据滤波的基本思想，针对杨高印提出的小子域滤波法存在子域划分重心不稳的问题，本文提出了新的子域划分方法，即通过考察滤波窗口内数据的数值分布或梯度变化情况，在充分考虑子域划分对称性的基础上，采用对称子域的均方根误差或平均梯度大小差值为判别指标确定数据变化的主梯度方向，并用此判别指标代替数据变化平缓系数判别指标，在滤波输出的取值上考虑数据变化的梯度方向上的两个相对称子域对输出的贡献大小进行加权平均，使滤波输出结果更稳定。通过数据试算，表明网格数据滤波法输出结果更加合理、稳健。     

Bag filters for TPP

Cleaning of TPP flue gases with bag filters capable of pulsed regeneration is examined. A new filtering element with a three-dimensional filtering material formed from a needle-broached cloth in which the filtration area, as compared with a conventional smooth bag, is increased by more than two times, is proposed. The design of a new FRMI type of modular filter is also proposed. A standard series of FRMI filters with a filtration area ranging from 800 to 16,000 m² is designed for an output more than 1 million m₃/h of with respect to cleaned gas. The new bag filter permits dry collection of sulfur oxides from waste gases at TPP operating on high-sulfur coals. The design of the filter makes it possible to replace filter elements without taking the entire unit out of service. __________ Translated from élektricheskie Stantsii, No. 3, March 2007, pp. 43–46.     

Nonlinear filtering structure for image smoothing in mixed-noise environments

This paper introduces a new nonlinear filtering structure for filtering image data that have been corrupted by both impulsive and nonimpulsive additive noise. Like other nonlinear filters, the proposed filtering structure uses order-statistic operations to remove the effects of the impulsive noise. Unlike other filters, however, nonimpulsive noise is smoothed by using a maximum a posteriori estimation criterion. The prior model for the image is a novel Markov random-field model that models image edges so that they are accurately estimated while additive Gaussian noise is smoothed. The Markov random-field-based prior is chosen such that the filter has desirable analytical and computational properties. The estimate of the signal value is obtained at the unique minimum of the a posteriori log likelihood function. This function is convex so that the output of the filter can be easily computed by using either digital or analog computational methods. The effects of the various parameters of the model will be discussed, and the choice of the predetection order statistic filter will also be examined. Example outputs under various noise conditions will be given.