视频图像中的实时人脸检测方法

给出了一种视频图像中的实时人脸检测方法，该方法综合了彩色视频图像的运动信息和颜色信息，可以快速地对图像中的人脸区域进行定位。算法通过对视频图像序列中每连续3帧图像进行对称差分，提取出运动区域；然后基于肤色聚类模型，再对运动区域进行肤色检测，经过候选人脸验证，最终定位图像中的人脸。实验表明，提出的方法检测速度快，实现简单、高效，满足实时系统的要求。     

Minimum perimeter coverage of query regions in a heterogeneous wireless sensor network

Many applications of wireless sensor networks (WSNs) require the tracking and the surveillance of target objects. The information to be gathered about the target may be related to only a particular monitored sub-area. Therefore, it is desirable to direct queries only to the nodes monitoring this particular area and for energy saving reasons, only the queried nodes should respond. In this paper, we introduce two new efficient distributed algorithms for finding the minimum number of connected perimeter sensor nodes that are sufficient to cover the perimeter of queried region, where the union of their sensing regions covers the perimeter of queried area. The simulation results show that our proposed algorithms achieve significant reduction in communication load, while preserving full perimeter coverage.     

Two and three-dimensional intrusion object detection under randomized scheduling algorithms in sensor networks

We are interested in wireless sensor networks which are used to detect intrusion objects such as enemy tanks, cars, submarines, etc. Since sensor nodes have a limited energy supply, sensor networks are configured to put some sensor nodes in sleep mode to save energy. This is a special case of a randomized scheduling algorithm. Ignored by many studies, an intrusion object’s size and shape are important factors that greatly affect the performance of sensor networks. For example, an extremely large object in a small sensor field can easily be detected by even one sensor node, no matter where the sensor node is deployed. The larger an intrusion object is, the fewer sensor nodes that are required for detection. Furthermore, using fewer sensor nodes can save resources and reduce the waste of dead sensor nodes in the environment. Therefore, studying coverage based on intrusion object’s size is important. In this paper, we study the performance of the randomized scheduling algorithm via both analysis and simulation in terms of intrusion coverage intensity. In particular, we study cases where intrusion objects occupy areas in a two-dimensional plane and where intrusion objects occupy areas in a three-dimensional space, respectively. We also study the deployment of sensor nodes when intrusion objects are of different sizes and shapes. First, sensor nodes are deployed in a two-dimensional plane and a three-dimensional space with uniform distributions. Then, they are deployed in a two-dimensional plane and a three-dimensional space in two-dimensional and three-dimensional Gaussian distributions, respectively. Therefore, our study not only demonstrates the impact of the size and shape of intrusion objects on the performance of sensor networks, but also provides a guideline on how to configure sensor networks to meet a certain detecting capability in more realistic situations.     

Voxel-based surface area estimation: from theory to practice

Consider a complex, highly convoluted three-dimensional object that has been digitized and is available as a set of voxels. How can one estimate the (original, continuous) area of a region of interest on the surface of the object? The problem appears in the analysis of segmented MRI brain data and in other three-dimensional imaging applications. Several difficulties arise. First, due to the complexity of the surface and its foldings, the region of interest and its intended boundary can be concealed and are therefore difficult to delineate. Second, the correct surface topology on intricate voxel sets may not be obvious. Third, the surface area of a digital voxel world is generally very different than the area of the underlying continuous surface. These problems can be partly circumvented by transforming the voxel data to a polyhedral surface representation. Our challenge is to accomplish the task while maintaining the original voxel representation. Estimators for the continuous surface area of digital objects have been available for some time. However, the known methods are limited to fairly smooth and “well-behaved” surfaces. This research bridges the gap between the available surface area estimation theory, that applies to idealized settings, and the reality of MRI brain data. Surface connectivity ambiguities are alleviated by considering the object/background boundary voxel faces rather than the border voxels themselves. The region of interest on the surface is delimited by growing geodesic paths between user-provided anchor points. Surface estimation is extended to admit surfaces with higher curvature than previously considered. Performance evaluation results are provided, and operation on MRI brain data is demonstrated.     

Detecting faces in images: a survey

Images containing faces are essential to intelligent vision-based human-computer interaction, and research efforts in face processing include face recognition, face tracking, pose estimation and expression recognition. However, many reported methods assume that the faces in an image or an image sequence have been identified and localized. To build fully automated systems that analyze the information contained in face images, robust and efficient face detection algorithms are required. Given a single image, the goal of face detection is to identify all image regions which contain a face, regardless of its 3D position, orientation and lighting conditions. Such a problem is challenging because faces are non-rigid and have a high degree of variability in size, shape, color and texture. Numerous techniques have been developed to detect faces in a single image, and the purpose of this paper is to categorize and evaluate these algorithms. We also discuss relevant issues such as data collection, evaluation metrics and benchmarking. After analyzing these algorithms and identifying their limitations, we conclude with several promising directions for future research     

Secure data aggregation in wireless sensor networks: A watermark based authentication supportive approach

In-network processing presents a critical challenge for data authentication in wireless sensor networks (WSNs). Current schemes relying on Message Authentication Code (MAC) cannot provide natural support for this operation since even a slight modification to the data invalidates the MAC. Although some recent works propose using privacy homomorphism to support in-network processing, they can only work for some specific query-based aggregation functions, e.g. SUM, average, etc. In this paper, based on digital watermarking, we propose an end-to-end, statistical approach for data authentication that provides inherent support for in-network processing. In this scheme, authentication information is modulated as watermark and superposed on the sensory data at the sensor nodes. The watermarked data can be aggregated by the intermediate nodes without incurring any en route checking. Upon reception of the sensory data, the data sink is able to authenticate the data by validating the watermark, thereby detecting whether the data has been illegitimately altered. In this way, the aggregation–survivable authentication information is only added at the sources and checked by the data sink, without any involvement of intermediate nodes. Furthermore, the simple operation of watermark embedding and complex operation of watermark detection provide a natural solution of function partitioning between the resource limited sensor nodes and the resource abundant data sink. In addition, the watermark can be embedded in both spatial and temporal domains to provide the flexibility between the detection time and detection granularity. The simulation results show that the proposed scheme can successfully authenticate the sensory data with high confidence.     

Information extraction from sensor networks using the Watershed transform algorithm

Wireless sensor networks are an effective tool to provide fine resolution monitoring of the physical environment. Sensors generate continuous streams of data, which leads to several computational challenges. As sensor nodes become increasingly active devices, with more processing and communication resources, various methods of distributed data processing and sharing become feasible. The challenge is to extract information from the gathered sensory data with a specified level of accuracy in a timely and power-efficient approach. This paper presents a new solution to distributed information extraction that makes use of the morphological Watershed algorithm. The Watershed algorithm dynamically groups sensor nodes into homogeneous network segments with respect to their topological relationships and their sensing-states. This setting allows network programmers to manipulate groups of spatially distributed data streams instead of individual nodes. This is achieved by using network segments as programming abstractions on which various query processes can be executed. Aiming at this purpose, we present a reformulation of the global Watershed algorithm. The modified Watershed algorithm is fully asynchronous, where sensor nodes can autonomously process their local data in parallel and in collaboration with neighbouring nodes. Experimental evaluation shows that the presented solution is able to considerably reduce query resolution cost without scarifying the quality of the returned results. When compared to similar purpose schemes, such as “Logical Neighborhood”, the proposed approach reduces the total query resolution overhead by up to 57.5%, reduces the number of nodes involved in query resolution by up to 59%, and reduces the setup convergence time by up to 65.1%.     

Efficient Data Reporting in Intelligent Transportation Systems

Recently, intelligent transportation systems (ITSs) have emerged. These systems can improve traditional transportation systems and provide traffic information to travelers. In the area of transportation, wireless sensor networks (WSNs) can replace the existing wired sensors and expensive traffic monitoring systems to mitigate the time and costs of installing such systems. However, accurate and on-time traffic information delivery is a major challenge, considering the energy constraints of sensor nodes. In this paper, we propose a two-tier architecture that includes a network of mobile objects (vehicles) in the upper layer and a hierarchical WSN in the bottom layer. Using this approach, a portion of loads on the low-power static sensor nodes can be transferred to mobile objects, such as powerful mobile devices. Moreover, to provide accurate and timely traffic information, a QoS-aware link cost function has been proposed and used for data transmission between the static sensor nodes. In addition, due to the mobility of the objects and the probability of losing packets in the mobile object tier, a reliable data forwarding mechanism has been proposed for this tier. In this mechanism, data packets are forwarded to the neighbors, which enhance the probability of the packets’ being received. The performance evaluation results indicate the effectiveness of the proposed architecture and data reporting mechanism for use in ITS applications.     

Towards Spatial Window Queries Over Continuous Phenomena in Sensor Networks

Recent research on sensor networks has focused on the efficient processing of declarative SQL queries over sensor nodes. Users are often interested in querying an underlying continuous phenomenon such as a toxic plume, whereas only discrete readings of sensor nodes are available. Therefore, additional information estimation methods are necessary to process the sensor readings to generate the required query results. Most estimation methods are computationally intensive, even when computed in a traditional centralized setting. Furthermore, energy and communication constraints of sensor networks challenge the efficient application of established estimation methods in sensor networks. In this paper, we present an approach using Gaussian kernel estimation to process spatial window queries over continuous phenomena in sensor networks. The key contribution of our approach is the use of a small number of Hermite coefficients to approximate the Gaussian kernel function for subclustered sensor nodes. As a result, our algorithm reduces the size of messages transmitted in the network by logarithmic order, thus saving resources while still providing high-quality query results.     

TGM-COT: energy-efficient continuous object tracking scheme with two-layer grid model in wireless sensor networks

In a resource-constrained wireless sensor network, energy efficiency is a principle issue for monitoring the movement of continuous objects, such as wild fire and hazardous chemical material. In this paper, a continuous object tracking scheme with two-layer grid model (TGM-COT) is proposed. To address the problem of boundary distortion caused by uneven node distribution, we put forward a novel mechanism for boundary nodes identification. Furthermore, a streamlining mechanism is designed to reduce the amount of uploaded data. Simulation results demonstrate that, without sacrificing additional energy consumption, TGM-COT is able to achieve high tracking accuracy and significantly reduce the communication overhead.     

数据集成技术在树型WSN中的应用与研究

针对无线传感器网络中节点能量有限这一特点,如何提高能量利用率、延长网络寿命是每个WSN研究者所必须面临的问题,同样也是WSN发展过程中必须被解决的难题。为此,本文将一种数据集成算法引入无线传感器网络,通过去除节点间的冗余信息来降低网络中的数据流量,从而降低网络能耗,达到提高网络能量利用率和延长网络寿命的目的。文章以树型无线传感器网络为例,分别对采用数据集成算法前后的网络能量消耗进行分析研究,并给出了相应的能量消耗模型。随后又通过计算机仿真和实验数据监测的方法,验证了此数据集成算法在降低树型WSN能耗中是有效的,而且具有一定的实用价值。     

闭坑矿井矿界煤柱采动损伤及其安全性评价

闭坑矿井矿界煤柱受采动损伤后其隔水能力降低,易导致闭坑老空水破坏矿界煤柱进入相邻生产矿井,影响矿井安全生产。以淮北矿区两相邻矿井即闭坑矿井沈庄矿和生产矿井袁庄矿的矿界煤柱为研究对象,采用FLAC 3d数值模拟和理论计算方法,对沈庄矿S2Ⅱ313工作面和袁庄矿Ⅲ3142工作面矿界煤柱采动损伤区宽度进行研究,根据研究结果确定矿界煤柱采动损伤区宽度为13.1 m,表明矿界煤柱已被破坏,两相邻矿井边界煤岩层已不具备阻水能力。在此基础上对矿界煤柱安全性进行了评价,得出现有矿界煤柱存在渗水和溃水危险性,需采取布置阻水帷幕或注浆加固等措施。     

Pattern recognition for detecting distributed node exhaustion attacks in wireless sensor networks

Malicious attacks when launched by the adversary-class against sensor nodes of a wireless sensor network, can disrupt routine operations of the network. The mission-critical nature of these networks signifies the need to protect sensory resources against all such attacks. Distributed node exhaustion attacks are such attacks that may be launched by the adversarial class from multiple ends of a wireless sensor network against a set of target sensor nodes. The intention of such attacks is the exhaustion of the victim’s limited energy resources. As a result of the attack, the incapacitated data-generating legitimate sensor nodes are replaced with malicious nodes that will involve in further malicious activity against sensory resources. One such activity is the generation of fictitious sensory data to misguide emergency response systems to mobilize unwanted contingency activity. In this paper, a model is proposed for such an attack based on network traffic flow. In addition, a distributed mechanism for detecting such attacks is also defined. Specific network topology-based patterns are defined to model normal network traffic flow, and to facilitate differentiation between legitimate traffic packets and anomalous attack traffic packets. The performance of the proposed attack detection scheme is evaluated through simulation experiments, in terms of the size of the sensor resource set required for participation in the detection process for achieving a desired level of attack detection accuracy. The results signify the need for distributed pattern recognition for detecting distributed node exhaustion attacks in a timely and accurate manner.     

基于三维地形修正的无线传感器网络覆盖盲区检测

三维地形下传感器节点经随机部署后,由于地形起伏的特点,会导致覆盖盲区的产生.为了能够有效检测覆盖盲区,本文提出了一种基于三维地形修正的无线传感器网络覆盖盲区检测方法,通过建立单位球感知模型,随机部署传感器节点,对节点进行Delaunay三角剖分.计算出三角形的外接圆,根据边界检测算法求出覆盖盲区边界,并且剔除假边界节点,得出改善后的边界.计算传感器节点的坡度和坡向信息,根据地形修正原理计算出传感器节点的实际探测半径,最终得出覆盖盲区的最小边界.仿真实验结果验证了该检测方法可有效检测三维地形下的覆盖盲区,对于起伏较大的地形也有一定的适应性.     

Collision Detection for Animation using Sphere-Trees

The detection of collisions between moving polyhedral objects is one of the most computationally intensive tasks in the computer animation process. The use of object-oriented techniques to encapsulate data within the objects' structures compounds this problem through the requirement for inter-object message passing in order to obtain geometric information for collision detection. The REALISM system decreases the time for collision detection by using a three stage process. The first stage identifies objects in the same locality using a global bounding volume table. The second stage locates regions of possible collision using a sphere-tree data structure (a hierarchical tree of spheres based on octree-type spatial subdivision). The final stage finds intersections between polygonal faces of the objects that are contained within the intersecting pairs of leaf nodes. Hence the algorithm uses a spherical geometry approximation rapidly to locate regions of potential collisions and then uses a local intersection test with actual object geometry information. The system is therefore fast and accurate. Tests for various geometric objects support this and show performance improvements of jive times over traditional polyhedral intersection tests.     

Obstacle detection and estimation in wireless sensor networks

In wireless sensor networks (WSNs), it is likely that a deployed area contains obstacles of some form. These obstacles may potentially degrade the functionality of the WSN. If the size and location of the obstacles can be detected, their influence can be reduced. Accordingly, this paper describes a scheme for detecting obstacles in WSNs. The scheme identifies the obstacles by marking the sensor nodes around the obstacle boundaries. The scheme does not require the absolute position of individual nodes in the sensing field nor any additional hardware, and thus can significantly reduce the deployment costs. The efficiency of the scheme is demonstrated via simulations performed using the network simulator ns-2. The results show that the detection scheme needs much less overhead compared to previous research while still marking the nodes close to the obstacles precisely.     

Coverage area enhancement in wireless sensor network

In the wireless sensor network, coverage area may be enhanced after an initial deployment of sensors. Though, some research works propose how to decrease the coverage hole by increasing sensing range or movement assisted sensor deployment, these are not suitable for energy constraint wireless sensor network, as longer mobility distance or higher power level consume more energy. In this paper, we address the increasing coverage area through smaller mobility of nodes. We find out the coverage hole in the monitoring region, which is not covering by any sensing disk of sensor. Then, we address the new position of mobility nodes to increase the coverage area. The simulation result shows the mobile nodes can recover the coverage hole perfectly. The coverage holes is recovered by mobility on the existing recovery area, which cannot be lost. Moreover, hole detection time in our proposed protocol is better than existing algorithm.

     

Precise glasses detection algorithm for face with in-plane rotation

This paper proposes an accurate glasses detection algorithm for in-plane rotated faces. Firstly, to normalize the face with in-plane rotation to be an upright one, this paper proposes a new set of ±26.565° haar-like features which can be calculated quickly to represent the features of rotated faces. With these new features, we address to build different rotated detectors by rotating upright face detector directly so as to achieve in-plane rotated face detection. Secondly, according to the rule that the objects with fixed and complex structure in a simple background are much easily to be detected, as the eyes have much complex structure in the face, we propose to detect the eyes for minimizing the glasses searching region. Because the nosepiece region of the glasses has much fixed structure and simple background, we use this part to train a new detector with the haar-like features and AdaBoost algorithm for precise glasses detection. This approach accomplishes glasses detection for in-plane rotated face with high speed and precision. The experimental results on the CMU-MIT, BioID, and CAS-PEAL-R1 face databases demonstrate that the algorithms for face, eye and glasses detection yield accurate results and maintain the speed advantage.     

基于线性回归的无线传感器网络分布式数据采集优化策略

事件监测是无线传感器网络中最重要的应用之一,部署在监测区域内的传感器节点通过对感知数据信息的采集、处理和传输等基本操作完成具体的监测任务,在各种操作中,节点之间的数据传输是最消耗能量的.为了减少节点之间的通信数据量,达到降低网络能耗和延长网络生命周期的目的,该文提出了一种能量高效的基于线性回归的无线传感器网络分布式数据采集优化策略,通过应用线性回归分析方法构建感知数据模型,保持感知数据的特征,使节点仅传输回归模型的参数信息,代替传输实际监测的感知数据信息.仿真实验结果表明,文中提出的数据采集优化策略能通过较小的通信量有效地实现事件监测区域感知数据的预测和估计,降低网络的总能量消耗,延长网络的生命周期.