查询词典 edge detection

In this paper, the theory, algorithm, and experiment of automatic object detection and tracking are studied in depth. It is firstly pointed out that the essential of Mean Shift method is a special Newton-Gaussian method. A new method named Fast Mean Shift is established to stretch the conservative step of Mean Shift method. The convergence and validity of this new method are proved in theory. And it is also proved that the convergence speed of Fast Mean Shift is faster than that of Mean Shift. The contrast experiments of searching the maximum possibility density of random of data sets in plane and 3D space are done. The results show that this new method can reduce the iterations greatly. A new object tracking method based on Fast Mean Shift is built to improve the object tracking performance, which is shown in the face tracking experiment with the tennis sequence provided by the Ohio State University, and the car tracking experiment with the car sequence provided by Kalsruhe University. The face trcking experiment with highly noised images show that the object tracking method based on Fast Mean Shift has strong anti-jamming ability. A new fast color object detection technology based on characteristic color is established, which use characteristic color distribution to compute the characteristic color vector of any area in an image quickly. With the high performance search method, the fast object detection is achieved. At last, using object tracker based on Fast Mean Shift and color object detector based on characteristic color with the Kalman filter, PID controller, searial communication and other technologies, automatic object detection and tracking system with control system is built. The availability and anti-jamming ability of this system are verified by some object detection and tracking tests in different scenes.

本文对目标自动识别与跟踪进行了理论、计算、试验三方面的深入研究，主要包括：首次指出了目标跟踪技术中常用的均值迁移方法的本质为一种特殊的高斯-牛顿方法，改进了均值迁移方法步长取值保守的弱点，建立了快速均值迁移方法，证明了该方法的收敛性、有效性以及收敛速度优于均值迁移方法；进行了平面和3维随机分布数据集的最大概率密度搜索对比试验，试验结果表明，快速均值迁移方法大大减少了迭代次数；建立了基于快速均值迁移的目标跟踪方法，利用俄亥俄州立大学提供的乒乓球序列图像和卡斯鲁厄大学的汽车序列图像，对人脸和汽车目标跟踪性能分别进行了对比试验，并进行了高噪声人脸图像目标跟踪试验，结果表明，基于快速均值迁移的目标跟踪方法有效提高了目标跟踪性能，具有很强的抗干扰能力；建立了一种新型彩色目标自动识别方法，采用特征色彩分布函数实现了对任意图像区域特征色彩矢量的快速计算，建立了高效的搜索方法，实现了彩色目标的快速识别；将基于快速均值迁移方法的目标跟踪方法、基于特征色彩的目标识别方法与卡尔曼滤波、PID控制、串行通讯等技术结合，建立了带有控制系统的快速目标自动识别与跟踪系统，并在不同场景下进行了目标自动识别与跟踪试验，验证了快速目标自动识别与跟踪系统的有效性和抗干扰能力。

A series PCR amplification for differential control strains and DNA samples diluted gradient (1:10) have been used to evaluate the specificity and sensitivity of PCR assay established.Results 1. Detection of GAS by PCR assay: The 345bp specific fragment of speB gene were amplified in all the tested GAS strains including three strains of scarlet fever, whereas it was detected in none of the differential control strains. The lowest limit of detection was 6.5pg genome DNA of GAS strain. 2. Detection of corynebacterium diphtheria by PCR assay: The318bp specific fragment of toxB gene were amplified in all the tested toxigenic corynebacterium diphtheria strains, whereas it was detected in none of the differential control strains. The lowest limit of detection is 850fg/μl genome DNA of corynebacterium diphtheria strain. 3. Detection of Lp by PCR assay: The 340bp specific fragments of mip gene were amplified in all the tested Lp strains, whereas it was detected in none of the differential control strains including three strains of non-pneumophila.

结果：1、用PCR方法检测A组链球菌：以A组链球菌致热性外毒素基因speB为靶序列，设计的扩增引物对全部对照菌株的扩增结果为阴性，而全部A组链球菌参考株均能扩增出特异的345bp片段，其中包括三株猩红热链球菌，检测敏感性为6.5pg／μl DNA.2、用PCR方法检测白喉杆菌：以白喉外毒素基因toxB为靶序列，设计的扩增引物对全部白喉杆菌参考株均能扩增出特异的318bp片段，而全部对照株的扩增结果为阴性，检测敏感性为850fg／μl DNA.3、用PCR方法检测嗜肺军团菌：以嗜肺军团菌巨噬细胞感染增强子基因mip为靶基因，设计的引物对嗜肺军团菌14个血清型参考株均扩增出特异的340bp片段，而鉴别对照株包括三株非嗜肺军团菌均未扩增出任何片段。

With intrusion detection technology explored, and the cause of some important issues in intusion detection field that must be resolved immediately annalyzed, an Enhanced Intrusion Detection Agent ,, is presented, which works under the Agent-based Distributed Intrusion Detection framework, with an enhanced detection engine inside, and obtaining decision-making information from an Assistant Information Collecting Agent .

本文对入侵检测技术进行了分析和研究，通过分析入侵检测领域亟待解决的一些问题产生的根源，提出了在基于软件代理的分布式入侵检测模型框架下，使用一种以增强型入侵检测引擎为核心的入侵检测代理，并引入辅助的信息收集代理，协同解决基于模式匹配的网络入侵检测系统高虚警率低检测效率的设计思想，并在此想法的基础上实现了一个增强型的入侵检测代理(Enhanced Network Intrusion Detection Agent， E_NIDA)和一个辅助的信息收集代理(Assistant Information Collecting Agent，。A_ICA)。

Along with enterprise paying more attention to the importance of quality detection and detection instruments using popularly,a lot of information is stored, so it is urgent to build dealing with detective data information system. This system could get timely product quality data, and grasp the quality change trend in order to support effectively maintenance of old products and decision-making for new product development, as well as help the quality detection personnel extricating of repeated , burdensome statistic and handworked work,This paper firstly bring forward storing tabacco-quality-detection data by using database technology,aiming at data storing of product detection and further analyzing demand.

随着企业对质量检测重要性认识的不断加大，检测仪器的大量使用，许多企业已经积累了大量的检测数据，迫切需要一套储存、加工处理分析检测数据的信息系统，把质量检测人员从重复繁杂的统计计算和手工作业中解脱出来，获取及时的产品质量数据，实时把握产品质量变化趋势，以求更有效地支持老产品的维护和新产品开发的战略决策。

This thesis from the angle of signal detection, first theoretically gives out a RFOG closed signal detection design based on DSP. The signal detection system mainly includes related detection. signal conversion and the formation of feedback saw tooth. Related detection must guarantee signal synchronization sampling, and get the biggest signal noise ratio.

本文首先分析了谐振式光纤陀螺的基本原理和系统结构，并从信号检测的角度出发，从理论上给出了一个基于DSP实现的谐振式光纤陀螺的闭环信号检测方案，信号检测系统主要包括相关检测，信号的转换和反馈锯齿波的形成。

Results: The PreCS intersected the cerebral longitudinal fissure in the shape of "Y", and the detection rate of MRI was 8.33%; the PreCS intersected the superior frontal sulcus, the middle frontal sulcus and the inferior frontal sulcus in the shape of inverted "Y", and the detection rates of MRI were 90.00%, 3.33%, and 76.67%, respectively; the PreCS was not only parallel to central sulcus, hut also intersected the CS, and the detection rates of MRI were 83.33% and 16.67%, respectively; the PreCS intersected the lateral fissure in the shape of inclined "Y", and the detection rate of MRI was 80.00%; the PreCS was paralleT to the anterior ascending ramus of lateral fissure, and the detection rate of MM was 98.33%.

结果：大脑中央前沟与大脑纵裂呈&Y&形相交，MRI显示率为8.33%；大脑中央前沟与额上沟、额中沟及额下沟均呈倒&T&形相交，MRT显示率分别为90.00%、3.33%和76.67%；大脑中央前沟与中央沟有近似乎行排列和呈倒&Y&形相交2种情形，MRI显示率分别为90.00%、83.33%和16.67%；大脑中央前沟与中央沟有近似平行排列和呈倒&Y&形相交，MRI显示率为80.00%；大脑中央前沟与外侧裂前升支呈平行排列，MRI显示率为98.33%。

VC prepared by an image processing procedures : it can achieve image edge detection and contour extraction, Edge Detection Operator : Robert operator, bobel operator, prewitt operator, kirsch operator, gauss - laplace operator. Hough line detection, contour extraction, seed testing and contour tracking.

用VC编写的一个图像处理的程序：它可以实现图像的边缘检测和轮廓提取，边缘检测算子包括：Robert算子、bobel算子、prewitt算子、kirsch算子、gauss－laplace算子。hough直线检测、轮廓提取、种子检测和轮廓跟踪。

At last, the application of the algorithm proposed to image edge detection and comparative experiments show that the algorithm have great ability of detection the fuzzy edge and exiguous edge.

最后将该算法应用到图像边缘检测，对比实验表明，该算法具有很强的模糊边缘和微细边缘检测能力。

The ordinary edge detection methods such as Sobel and Canny operators are immune to noise, which result in cophasal axes discontinuity due to lack of edge information caused by noise. In order to extract cophasal axes more continuous and immune to noise, a novel approach is developed based on edge detection. Firstly, the median filter and improved Sobel method are used to preprocess seismic section images. Secondly, the threshold value is determined automatically based on variance to detect abnormal information. Finally, the complete cophasal axes are extracted after a post-processing.

由于边缘检测中常用的Sobel算子和Canny算子检测同相轴时对噪声免疫差，常因噪声因素引起边缘信息缺失导致同相轴提取不连续，本文采用中值滤波并对Sobel算子进行改进，对地震剖面图进行预处理，结合均方差自动确定阈值并进行异常检测，提取完整同相轴。

This paper gives a method of image edge detection,using multiple edge features and fuzzy reasoning.To effectively extract edges,edge pixels are classified according to the difference of intensity distribution,and multiple effective edge features are defined.

图像边缘检测的传统方法一般仅基于边缘邻域的一阶或二阶导数这一单一特征，并且对模糊边缘图像不敏感，针对此问题，本文提出基于多特征和模糊推理的边缘检测算法。

It has to be computed by the host for the actual clock rate.

主控制器可以用来计算实际的时钟速率。

Let's see what you got.

让我看看你的东西