Object segmentation and labeling by learning from examples

We propose a system that employs low-level image segmentation followed by color and two-dimensional (2-D) shape matching to automatically group those low-level segments into objects based on their similarity to a set of example object templates presented by the user. A hierarchical content tree data structure is used for each database image to store matching combinations of low-level regions as objects. The system automatically initializes the content tree with only "elementary nodes" representing homogeneous low-level regions. The "learning" phase refers to labeling of combinations of low-level regions that have resulted in successful color and/or 2-D shape matches with the example template(s). These combinations are labeled as "object nodes" in the hierarchical content tree. Once learning is performed, the speed of second-time retrieval of learned objects in the database increases significantly. The learning step can be performed off-line provided that example objects are given in the form of user interest profiles. Experimental results are presented to demonstrate the effectiveness of the proposed system with hierarchical content tree representation and learning by color and 2-D shape matching on collections of car and face images.     

结构化编程实现Sniffer

随着各高校校园网络规模的不断扩大和各种应用的急剧增加,各种新的网络问题也在不断出现。Sniffer是网络管理员进行网络管理的有效工具,可以有效地对网络进行实时监控,及时发现网络中存在的各种问题。本文首先介绍了Sniffer的工作原理,其次给出了实现Sniffer功能的编程流程图,并用C语言给出了编程所需的关键代码。     

数值试井在白6-2井测试资料分析评价中的应用

数值试井是在近来发展的一项新的试井解释技术，具有处理非均质，复杂边界油气藏问题的优点，弥补了常规解析试井技术的不足。运用数值试井解释技术，结合动、静态资料，通过对白6-1断块白6-2井测试资料的综合分析评价，确定了储层渗流参数，落实了该气藏的边界情况，为该断块下一步开发提供可靠依据。     

SMA沥青混合料在皖北地区高速公路中的应用

皖北地区夏季高温多雨,冬季严寒,气候较为恶劣,在该地区宜推广应用SMA混合料铺筑表面层。在进行SMA配合比设计时,必须严格控制原材料的质量,经过精心设计、精心施工,才能修建出一条成功的SMA路面。     

弹性-粘弹性复合结构随机响应的各阶谱矩的计算方法

提出了一种新的计算弹性-粘弹性复合结构随机响应的各阶谱矩的计算方法，它是一种时域复模态分析方法。利用此方法获得了弹性-粘弹性复合结构在白噪声、滤过白噪声等典型平稳随机激励下随机响应的各阶谱矩的解析表达式，分析了粘弹性对各阶谱矩的影响。此计算方法简便、易用，无论单自由度或多自由度系统均适用，为进一步研究弹性-粘弹性复合结构在随机激励下的可靠性打下良好基础。     

SCB火工品的研究与发展

文章评述了半导体桥火工品问世以来的研究成果与发展趋势。主要内容有半导体桥作用机理、半导体桥的结构与封装、半导体桥火工品的特点、半导体桥火工品的应用、半导体桥火工品的研究和发展趋势。     

丙烯醛制备1,3-丙二醇的中型试验研究

对丙烯醛水合、加氢制1，3-丙二醇工艺进行了中试研究，主要对小试技术的可靠性进行验证。结果表明，丙烯醛经水合、加氢制1，3-丙二醇的工艺技术可行，反应器、精馏塔等设备设计合理。中试丙烯醛水合单程转化率为45％～55％，3-羟基丙醛选择性达到了93％左右，3-羟基丙醛加氢收率接近100％，产品1，3-丙二醇的醛基质量分数降至5μg／g以下。与小试结果相比较，中试的水合选择性明显提高，加氢催化剂活性寿命延长。     

网络管理中事件关联检测机制的研究

网络事件的关联检测是网络管理需要解决的一个关键问题,本文根据分布式网络管理的特点,首次明确定义了网络事件的基本关联关系,在此基础上提出了一种基于Petri Net的事件关联检测机制,实现了基于事件内容的细粒度关联检测,并且在事件检测中充分考虑了时间因素,有效地提高了事件关联检测的准确性。     

USERS' TWO-DIMENSIONAL INTEREST AND ITS STORAGE METHOD

The users' interest can be mined from the web cache and can be used widely. The interest can be specialized by the two-tuple (term, weight) in the simple interest model, in which the association relations are not mined, and then the interest cannot be associated in expressing the users' interest. Based on analyzing the WWW cache model, this letter brings forward a two-dimensional interest model and gives the interrelated methods on how to store the two-dimensional interest model effectively.     

Nonlinear protein binding and enzyme heterogeneity: effects on hepatic drug removal

The kinetics of substrate removal by the liver and the resulting nonlinear changes in unbound fraction along the flow path at varying input drug concentrations were examined by a model simulation study. Specifically, we varied the binding association constant, KA, and the Michaelis-Menten constants (Km and Vmax) to examine the steady state drug removal (expressed as hepatic extraction ratio E) and changes in drug binding for (i) unienzyme systems and (ii) simple, parallel metabolic pathways; zonal metabolic heterogeneity was also added as a variable. At low KA, E declined with increasing input drug concentration, due primarily to saturation of enzymes; only small differences in binding were present across the liver. At high KA, a parabolic profile for E with concentration was observed; changes in unbound fraction between the inlet and the outlet of the liver followed in parallel fashion. Protein binding was the rate-determining step at low input drug concentrations, whereas enzyme saturation was the rate-controlling factor at high input drug concentration. Heterogeneous enzymic distribution modulated changes in unbound fraction within the liver and at the outlet. Despite marked changes in unbound fraction occurring within the liver for different enzymic distributions, the overall transhepatic differences were relatively small. We then investigated the logarithmic average unbound concentration and the length averaged concentration as estimates of substrate concentration in liver in the presence of nonlinear drug binding. Fitting of simulated data, with and without assigned random error (10%), to the Michaelis-Menten equation was performed; fitting was repeated for simulated data obtained with presence of a specific inhibitor of the high-affinity, anteriorly distributed pathway. Results were similar for both concentration terms: accurate estimates were obtained for anterior, high affinity pathways; an overestimation of parameters was observed for the lower affinity posteriorly distributed pathways. Improved estimations were found for posteriorly distributed pathways upon inhibition with specific inhibitors; with added random error, however, the improvement was much decreased. We applied the method for fitting of several sets of metabolic data obtained from rat liver perfusion studies performed with salicylamide (SAM) (i) without and (ii) with the presence of 2,6-dichloro-4-nitrophenol (DCNP), a SAM sulfation inhibitor. The fitted results showed that SAM sulfation was a high-affinity high-capacity pathway; SAM glucuronidation was of lower affinity but comparable capacity as the sulfation pathway, whereas SAM hydroxylation was of lower affinity and lower capacity.