Towards Collaborative Robotics in Top View Surveillance:A Framework for Multiple Object Tracking by Detection Using Deep Learning

Collaborative Robotics is one of the high-interest research topics in the area of academia and industry.It has been progressively utilized in numerous applications,particularly in intelligent surveillance systems.It allows the deployment of smart cameras or optical sensors with computer vision techniques,which may serve in several object detection and tracking tasks.These tasks have been considered challenging and high-level perceptual problems,frequently dominated by relative information about the environment,where main concerns such as occlusion,illumination,background,object deformation,and object class variations are commonplace.In order to show the importance of top view surveillance,a collaborative robotics framework has been presented.It can assist in the detection and tracking of multiple objects in top view surveillance.The framework consists of a smart robotic camera embedded with the visual processing unit.The existing pre-trained deep learning models named SSD and YOLO has been adopted for object detection and localization.The detection models are further combined with different tracking algorithms,including GOTURN,MEDIANFLOW,TLD,KCF,MIL,and BOOSTING.These algorithms,along with detection models,help to track and predict the trajectories of detected objects.The pre-trained models are employed;therefore,the generalization performance is also investigated through testing the models on various sequences of top view data set.The detection models achieved maximum True Detection Rate 93%to 90%with a maximum 0.6%False Detection Rate.The tracking results of different algorithms are nearly identical,with tracking accuracy ranging from 90%to 94%.Furthermore,a discussion has been carried out on output results along with future guidelines.

Incremental View Computation Model for Object-Oriented Information

We introduce a model to implement incremental update of views. The principle is that unless a view is accessed, the modification related to the view is not computed. This modification information is used only when views are updated. Modification information is embodied in the classes (including inheritance classes and nesting classes) that derive the view. We establish a modify list consisted of tuples (one tuple for each view which is related to the class) to implement view update. A method is used to keep views from re-update. Key words object-oriented database - incremental computation - view-computation - engineering information system CLC number TP 391 Foundation item: Supported by the National Natural Science Foundation of China(60235025)Biography: Guo Hai-ying (1971-), female, Ph. D, research direction: CAD and engineering information system.

融合光流与多视角几何的动态视觉SLAM系统

视觉同步定位与地图构建(SLAM)在动态干扰的情况下,导致定位精度下降且无法准确构建静态地图,提出一种结合光流和多视角几何的动态视觉SLAM系统,该系统是在ORB-SLAM2的基础上进行改进的。在追踪线程中引入处理后的光流信息,结合多视图几何,得到动态区域掩码对视野内图像帧进行分割,实现动态区域检测并滤除动态区域中的特征点,在保证视觉SLAM系统实时性的同时提高追踪准确度,替换原本的地图构建线程。在新的地图构建线程中,引入光流信息及Mobile NetV2实例分割网络。利用实例分割网络分割结果结合光流动态区域掩码对获取到的有序点云逐层分割,解决地图构建中动态物体造成的“拖影”问题。同时对分割后的点云团融合语义信息,最终构建静态语义八叉树地图。在TUM Dynamic Objects数据集上的实验结果表明,相较于ORB-SLAM2,在高动态场景序列测试中,该算法的定位精度平均提升70.4%,最高可提升90%。

视景仿真编辑器ViewEditor的设计与实现

我们开发的ＶｉｅｗＥｄｉｔｏｒ是一个基于ＷｉｎｄｏｗｓＮＴ系统的视景仿真编辑器。它集图形、图象、声音、文字编辑于一体，支持开放式接口，具有界面优良、操作直观灵活。满足ＷＹＳＩＷＹＣ等特点。该文首先介绍了ＶｉｅｗＥｄｉｔｏｒ的功能和系统构成，然后对ＶｉｅｗＥｄｉｔｏｒ的层次模型与界面设计进行了详细的论述。

基于GE View技术的虚拟机械臂对象设计

机械臂是机器人技术在生产领域中应用最广泛的自动化机械装置,其特点是能精确定位于二维或三维空间进行作业,其控制方式根据要实现的功能具有很大差别。因此,采用计算机技术实现虚拟的机械臂对象对于实践控制功能具有重要意义。通过阐述虚拟机械臂对象的设计步骤,介绍了利用GE View技术设计虚拟机械臂的过程和要点。分析了机械臂对象的功能定义、界面设计方法、信号和数据设计、正常和非正常控制下的动画设计。

顾及动态物体感知的增强型视觉SLAM系统

传统的同步定位与制图(Simultaneous localization and mapping,SLAM)系统在复杂环境下工作时,无法分辨环境中的物体是否存在运动状态,图像中运动的物体可能导致特征关联错误,引起定位的不准确和地图构建的偏差。为了提高SLAM系统在动态环境下的鲁棒性和可靠性,本文提出了一种顾及动态物体感知的增强型视觉SLAM系统。首先,使用深度学习网络对每一帧图像的动态物体进行初始检测,然后使用多视图几何方法更加精细地判断目标检测无法确定的动态物体区域。通过剔除属于动态物体上的特征跟踪点,提高系统的鲁棒性。本文方法在公共数据集TUM和KITTI上进行了测试,结果表明在动态场景中定位结果的准确度有了明显提升,尤其在高动态序列中相对于原始算法的精度提升在92%以上。与其他顾及动态场景的SLAM系统相比,本文方法在保持精度优势的同时,提高了运行结果的稳定性和时间效率。

用于MEMS振镜激光扫描的显微物镜设计

为满足微机电系统(micro-electro-mechanical system,MEMS)二维激光扫描系统对显微物镜小入瞳直径、大入射角度、大视场的要求,利用光学设计软件Zemax设计了一款入瞳直径为1.1 mm,可匹配MEMS二维振镜±18°大扫描角度的近红外无限共轭微型显微物镜。该物镜总长小于23 mm,数值孔径为0.4,分辨率为1.26μm,工作距为900μm,各项像差校正良好,满足使用需求。设计结果表明,该微型显微物镜可满足便携式皮肤检测仪器的MEMS二维振镜激光扫描系统的应用要求。

Point Cloud Processing Methods for 3D Point Cloud Detection Tasks

Light detection and ranging(LiDAR)sensors play a vital role in acquiring 3D point cloud data and extracting valuable information about objects for tasks such as autonomous driving,robotics,and virtual reality(VR).However,the sparse and disordered nature of the 3D point cloud poses significant challenges to feature extraction.Overcoming limitations is critical for 3D point cloud processing.3D point cloud object detection is a very challenging and crucial task,in which point cloud processing and feature extraction methods play a crucial role and have a significant impact on subsequent object detection performance.In this overview of outstanding work in object detection from the 3D point cloud,we specifically focus on summarizing methods employed in 3D point cloud processing.We introduce the way point clouds are processed in classical 3D object detection algorithms,and their improvements to solve the problems existing in point cloud processing.Different voxelization methods and point cloud sampling strategies will influence the extracted features,thereby impacting the final detection performance.

动态环境下基于深度学习的视觉SLAM研究

由于传统的同步定位与建图(simultaneous localization and mapping,SLAM)中有很强的静态刚性假设,故系统定位精度和鲁棒性容易受到环境中动态对象的干扰。针对这种现象,提出一种在室内动态环境下基于深度学习的视觉SLAM算法。基于ORB-SLAM2进行改进,在SLAM前端加入多视角几何,并与YOLOv5s目标检测算法进行融合,最后对处理后的静态特征点进行帧间匹配。实验使用TUM数据集进行测试,结果显示:SLAM算法结合多视角几何、目标检测后,系统的绝对位姿估计精度在高动态环境中相较于ORB-SLAM2有明显提高。与其他SLAM算法的定位精度相比,改进算法仍有不同程度的改善。

16世纪中国博古图与西方静物画中的物比较

中国博古图与西方静物画的兴盛都始于16世纪前后,都以静物为表现对象,但又有所不同。从物的属性来看,博古图中的物与人们的精神需求紧密相关,其主要功能经历了从宋代的正经补史到明代的雅文化象征再到清代的吉祥文化寓意的变迁;在西方静物画发展史上具有节点意义的罗马静物画、荷兰静物画、夏尔丹静物画的繁荣都与特定时期物质生活的变化紧密相关。从物的审美来看,博古图因物喻志,以志为本,体现了中国传统诗学的比兴理论;静物画观物绘形,以物为本,使被忽视或熟视无睹的事物变得陌生、反常,体现了陌生化的审美追求。从物的感知来看,两者体现了中西文化中感物与观物两种不同的美学传统及其背后的感官倚重的不同。但由于艺术有其自身的特征和发展规律,所以两者的这些不同只是相对的。

基于鸟瞰图的多任务端到端3D目标检测方法

由于在自动驾驶过程中激光雷达扫描过程中受到空间限制,采集到的点云数据损失了过多的信息。在聚合表示方法的基础上,提出了基于鸟瞰图的多任务端到端3D目标检测方法。首先,利用纹理特征提取器与语义特征提取器对特征金字塔的高级特征层与低级特征层进行特征融合,获得扩展的特征金字塔层,使得自动驾驶过程中的中等目标(骑自行车的人)与小目标(行人)不再耦合在同一级别的特征金字塔层,提高了点云拓扑信息的区域细节性。其次,在损失函数中引入用于分类的Focal loss与用于回归的CIOU loss,改善了正负样本比例,对目标检测框进行了约束,使得目标框在回归过程中能够以更高精度收敛。实验结果表明,提出的方法在自动驾驶中小目标检测中具有更强的检测能力与更高的检测精度。

Efficient View-Based 3-D Object Retrieval via Hypergraph Learning

View-based 3-D object retrieval has become an emerging topic in recent years,especially with the fast development of visual content acquisition devices,such as mobile phones with cameras.Extensive research efforts have been dedicated to this task,while it is still difficult to measure the relevance between two objects with multiple views.In recent years,learning-based methods have been investigated in view-based 3-D object retrieval,such as graph-based learning.It is noted that the graph-based methods suffer from the high computational cost from the graph construction and the corresponding learning process.In this paper,we introduce a general framework to accelerate the learning-based view-based 3-D object matching in large scale data.Given a query object Q and one object O from a 3-D dataset D,the first step is to extract a small set of candidate relevant 3-D objects for object O.Then multiple hypergraphs can be constructed based on this small set of 3-D objects and the learning on the fused hypergraph is conducted to generate the relevance between Q and O,which can be further used in the retrieval procedure.Experiments demonstrate the effectiveness of the proposed framework.

View Creation for Queriesin Object Oriented Databases

A view in object oriented databases corresponds to virtual schemawith restructured generalization and decomposition hierarchies. Numbers of viewcreation methodologies have been proposed. A major drawback of existing method-ologies is that they do not maintain the closure property. That is, the result of aquery does not have the same semantics as embodied in the object oriented datamodel. Therefore, this paper presents a view creation methodology that derives aclass in response to a user's query, integrates derived class in global schema (i.e.,considers the problem of classes moving in class hierarchy) and selects the requiredclasses from global schema to create the view for user's query. Novel idea of viewcreation includes: (a) an object algebra for class derivation and customization (wherethe derived classes in terms of object instances and procedure/methods are studied),(b) maintenance of closure property, and (c) classification algorithm which providesmechanism to deal with the problem of a class moving in a class hierarchy.

基于面向对象思想的SoS体系结构设计方法

针对一类由众多组件系统集结而成的系统之系统(system of systems,SoS),以美国国防部体系结构框架(DoDAF)为标准,提出了一种基于面向对象思想的SoS体系结构DoDAF作战视图产品五阶段迭代设计方法。利用UML静态和动态建模机制的特点,采用自顶向下、自底向上相结合的方式实现SoS体系结构作战视图产品的面向对象描述。以一个战术导弹防御(tactical missile defense,TMD)系统为例,详细说明SoS体系结构DoDAF作战视图产品的面向对象设计过程,并总结了该方法的两大优越特性:横向通用性与纵向可复用性,以及由此给SoS体系结构带来的"柔性"优势。这是传统设计方法所不能实现的。

基于Hibernate与Struts框架的数据持久化应用研究

基于使用单一框架构建企业级应用时存在拓展性差,结构复杂的问题,给出使用Hibernate和Struts两个开源框架进行整合开发的实例,同时探讨如何配置和灵活使用两种框架,简化对象持久化映射工作和开发中的MVC分工,以及如何充分发挥两者优势,构建结构清晰、具备强大扩展性和维护性的J2EE应用。

Object-oriented Integrated Multi-view Modeling Approach in CIMS

The consistency of the CIMS multi-view model is a complicated problem which plays an important role inthe CIMS life cycle. Based on the function-oriented multi-view modeling approach, an object-oriented integratedmulti-view modeling approach was further provided.CIM system is composed of several objects where each objecthas a multi-view description.The four most important view-points: function, information, resource and dynamic viewpoint were chosen to provide conceptual model of the object-oriented integrated multi-view modeling in CIMS using a system integration approach

基于MVC架构的LIBS软件设计与实现

为了方便使用实验室组装的激光诱导击穿光谱(LIBS)仪器,设计并实现了一套LIBS软件。该软件包括控制软件和光谱分析软件两部分,具有系统自检、数据采集、光谱预处理和元素识别等功能。采用MVC(Model-View-Controller)对软件的系统架构进行设计,描述了各Model的设计流程,并用C#语言完成了软件的开发。实验表明该软件可方便地操作仪器实现光谱采集和定性分析。

基于JSF框架的信息管理系统的设计和实现

JSF(Java server faces)是Java社区过程倡导的技术,是一个完整的Web解决方案。JSF是Sun制定的一个规范,未来很有可能包括在J2EE规范中。JSF以它所具备的条件使得应用程序开发变得简单、方便。在介绍了JSF的一般开发过程后,以目前流行的Eclipse为开发平台,后台加入实现数据封装的Hibernate技术,结合对象建设投资集团信息管理系统项目的开发生成一个功能简单的系统,并对JSF的应用做了探讨。

基于Struts框架的档案管理系统应用研究

详细介绍了MVC(模型-视图-控制)设计模式和Struts框架的工作原理,将基于MVC设计模式的Struts框架引入到档案管理系统的开发中,结合具体实例说明了如何应用Struts框架实现MVC设计模式,体现了软件框架结构和组件技术重用的优越性,提高了软件开发的效率,增强了软件的可扩展性和可维护性。

一种重叠视域多摄像机环境下的运动目标匹配SHFSM算法

针对重叠视域下多摄像机间目标匹配问题,提出了一种基于色度特征信息的色度特征区间集合匹配算法。通过分析目标颜色色度分量对光照强度的不变性,依据色度分量直方图定义了色度特征区间,并证明了其相关特性。该算法首先获取目标色度特征区间集合,并依据区间内特征信息像素的数量对特征区间进行排序,最后采用模糊匹配原理对排序的色度特征区间集合进行匹配。实验结果表明,算法匹配正确率可达95%,通过与经典算法对比体现了其准确率高、可靠性强的优点。