基于改进YOLOv3的运动目标分类检测算法研究

提出一种基于改进YOLOv3算法的一类运动目标检测算法. 为进一步提高YOLOv3的检测精度,采用基于DIoU优化的边界框回归损失函数进行计算; 优化非极大值抑制,有效减少了目标框重叠的现象,提高检测精度; 针对运动目标检测,提出一种基于目标框多中心点位移的检测算法. 经UA-DETRAC数据集上的实验表明,改进后的算法在提高检测精度的同时保证了较快的速度,准确率和召回率相比原始YOLOv3分别提高了 8.07%和3.87%,对运动目标的检测速度可达20 fps/s,可满足实时检测的要求.     

一起纵差保护跳闸开入异常分析与处理

针对一起500 kV线路故障进行分析,确定该线路分相电流差动保护跳闸位置开入异常,查出跳闸位置开入异常是由厂家配线错误导致的。本次异常处理对变电二次检修、基建验收等环节有一定的借鉴意义。     

有机涂层钢板生产线主要工艺

介绍了有机涂层钢板生产线上钢基板的预处理,全线张力控制,纠偏装置的布置,固化炉的选型.     

一种新的知觉分割模型*

根据知觉的最新理论研究和实验结果,提出一种新的知觉模型,并用于考查计算机视觉中的分割问题.分割问题实质上是知觉表达问题,现在分割的困难性可能是知觉上局部到整体的还原论假设造成的;分割可能不是一次完成的,而是经过整体到局部的过程才最终完成.新的知觉模型将对建立更好的分割模型提供理论基础.     

热型连铸单晶铜工艺参数对铸棒表面质量的影响

在自行设计，制造的横引式热型连铸设备上对影响连铸过程的各工艺参数进行试验，并就各工艺参数对铸件表面质量的影响机理进行了分析。结果表明，各工艺参数主要是通过影响固液界面在型口中的位置来影响铸棒的表面质量，只有对各工艺参数进行合理匹配，才能得到表面光滑的铸棒，此外，液位高度变化即液位波动，也会对铸棒的表面质量产生 一定影响。     

FMS实时控制系统动态建模方法(一)

在建立FMS物理对象的功能模型，状态转移图（STD），面向对象的Petri网（OPN）模型的基础上，本文将着色Petri网与面向对象技术相结合，提出了一种实时控制系统的建模方法－面向对象的着色Petri网（OCPN）建模方法，此建模方法在无锡职业学院的FMS的实时控制系统中得到验证。     

基于模型参考模糊自适应的电液位置伺服系统鲁棒控制研究

针对电液位置伺服系统存在的参数变化及外力扰动，引入模型参考模糊自适应机构，构成模型参考模糊自适应控制，即MRFAC，仿真实验表明，与传统PID相比，该方法可以有效抑制参数扰动及外力干扰，具有很强的鲁棒性。     

气动位置伺服系统的高精度控制研究

提出采用电-气压力比例阀构成气动位置伺服系统,运用全局线性化的观点线性化描述系统动态特性的数学方程,补偿了气缸活塞行程变化对系统控制性能的不良影响,并采用离散线性二次最优控制原理设计了系统的控制补偿器,实现了气动位置伺服系统的高精度控制。     

Outdoor attractancy of males ofMatsucoccus josephi (Homoptera: Matsucoccidae) andElatophilus hebraicus (Hemiptera: Anthocoridae) to synthetic female sex pheromone ofMatsucoccus josephi

The active component of the sex pheromone ofMatsucoccus josephi is (2E,6E,8E)-5,7-dimethyl-2,6,8-decatrien-4-one; the chemical is also a powerful kairomone of adult males and females of the bugElatophilus hebraicus the principal predator ofM. josephi. The presence of theZ isomer (2E,6Z,8E)-5,7-dimethyl-2,6,8-decatrien-4-one does not interfere with the attractancy of the activeE component forM. josephi males or the bug. Our results show a clear dose-response between trap catch ofM. josephi males andE. hebraicus. Conversely, increasing amounts of theZ isomer in the mixture did not affect the attraction of the scale insect males or the bug. The catch ofM. josephi males did not differ significantly among traps of different color, and was significantly higher with traps attached to the tree trunk than those suspended between trees. Comparison of the catch ofM. josephi among the three forests and between pine species suggests that the level of infestation ofPinus halepepsis andPinus brutia ssp.brutia is similar, despite the fact that the latter pine is resistant to the scale insect. Both sexes ofE. hebraicus were trapped in much lower numbers at the more infested sites. This may be related to interference with the activity ofE. hebraicus due to deterioration and drying of parts of the tree crowns and heavy colonization by generalist predators in injured trees.     

Online learning and fusion of orientation appearance models for robust rigid object tracking

We introduce a robust framework for learning and fusing of orientation appearance models based on both texture and depth information for rigid object tracking. Our framework fuses data obtained from a standard visual camera and dense depth maps obtained by low-cost consumer depth cameras such as the Kinect. To combine these two completely different modalities, we propose to use features that do not depend on the data representation: angles. More specifically, our framework combines image gradient orientations as extracted from intensity images with the directions of surface normals computed from dense depth fields. We propose to capture the correlations between the obtained orientation appearance models using a fusion approach motivated by the original Active Appearance Models (AAMs). To incorporate these features in a learning framework, we use a robust kernel based on the Euler representation of angles which does not require off-line training, and can be efficiently implemented online. The robustness of learning from orientation appearance models is presented both theoretically and experimentally in this work. This kernel enables us to cope with gross measurement errors, missing data as well as other typical problems such as illumination changes and occlusions. By combining the proposed models with a particle filter, the proposed framework was used for performing 2D plus 3D rigid object tracking, achieving robust performance in very difficult tracking scenarios including extreme pose variations.