红外激光车载云台去抖动设计

车辆行驶产生的颠簸、晃动,会严重影响红外激光车载云台采集视频的质量,不利于信息的观察和判读。为了改善视频质量,提出一种实时有效的车载电子稳像方法。首先,用非线性扩散滤波建立尺度空间,使视频帧在无损精度的前提下突出边缘信息;然后,采取图像亮度信息与梯度信息相结合的特征提取策略,利用图像亮度信息快速搜索潜在特征点并记录其位置,并通过分析其梯度信息与预判条件的对比结果来选取优质特征点,在特征描绘阶段,通过增强二值特征描绘器之间的显著性和差异性,提升特征量的辨别力;最后,通过位置验证辨别相似特征,提高全局运动矢量估计精度。时间比对实验表明提出的算法能够满足实时处理的要求,在分辨率为720 P时也能达到每秒处理帧频大于60帧;有效能力对比实验中,新方法的重复度均高于65%,明显提高了特征探测能力,并且在加入去抖动设计后6组实验的帧间转换精度分别提升了46.8%、30.8%、28.44%、28.1%、33.9%和53.4%,表明该方法极大地改善了抖动视频的稳定性和视频信息提取的准确度。     

Real-time image processing approach to measure traffic queueparameters

The real-time measurement of various traffic parameters including queue parameters is required in many traffic situations such as accident and congestion monitoring and adjusting the timings of the traffic lights. In case of the queue detection, at least two algorithms have been proposed by previous researchers. Those algorithms are used for queue detection and are unable to measure queue parameters. The authors propose a method based on applying the combination of noise insensitive and simple algorithms on a number of sub-profiles (a one-pixel-wide key-region) along the road. The proposed queue detection algorithm consists of motion detection and vehicle detection operations, both based on extracting edges of the scene, to reduce the effects of variation of lighting conditions. To reduce the computation time, the motion detection operation continuously operates on all the sub-profiles, but the vehicle detection is only applied to the tail of the queue. The proposed algorithms have been implemented on an 80386-based microcomputer system and the whole system works in real-time     

Simulation study on the interactivity of failure detection and signaling congestion control in intelligent optical networks

The Intelligent Optical Network (ION) has emerged as a promising technology in the development of optical networks. ION has a separate control plane, where real-time processing is a most substantial feature. However, the complicated control and management infrastructure requires multiple protocols to cooperate in the same signaling network, and jitter may impair real-time processing, like fast failure detection. In this article, the interactivity of failure detection and network congestion control will be analyzed and proposed solutions will be evaluated by simulations. The failure detection protocol in the control plane is inherited from IP networks. However, there is a scalability problem in the ION when requiring fast detection time. Congestion in the signaling network may be caused by various events. It could be fatal when interacting with fast failure detection would trigger a positive feedback. Several solutions will be presented and analyzed with respect to fast and scalable failure detection and congestion control. We performed a series of intensive simulations to study the interactivity of failure detection and congestion control on the 32-node USNET. The simulation results show that the proposed Enhanced Hello protocol performs well in both fast detection mode and congestion state, especially with far less false failure alarms. The Retransmission Reduction mechanism, which is designed to alleviate congestion, can avoid message loss in some degree, but gives little improvement in other aspects.     

人脸识别中视频回放假冒攻击的实时检测方法

利用合法用户的脸部视频进行回放假冒攻击,是目前人脸认证系统的重要安全威胁。针对此问题,本文提出了一种仅用单个普通摄像头来抵抗人脸视频假冒攻击的方法。不同于以往从人脸区域中获取假冒线索进行活体检测的方法,本文通过人脸输入图像与场景参考图像之间的背景对比,从人脸周围背景区域中寻求视频假冒攻击线索。首先,本文在尺度空间里构建人脸周围区域图像的背景特征点集合;然后,利用背景特征点集合建立识别场地背景和人脸背景的Gabor背景描述子,并用融合相位补差的相似度来进行背景比对。实验表明该方法能有效地识别视频回放假冒攻击。     

Robust energy-to-peak sideslip angle estimation with applications to ground vehicles

In this paper, the observer design problem for the sideslip angle of ground vehicles is investigated. The sideslip angle is an important signal for the vehicle lateral stability, which is not measurable by using an affordable physical sensor. Therefore, we aim to estimate the sideslip angle with the yaw rate measurements by employing the vehicle dynamics. The nonlinear lateral dynamics is modeled firstly. As the tyre model is nonlinear and the road adhesive coefficient is subject to a large variation, the nonlinear lateral dynamics is transformed into an uncertain model. Considering the variation of longitudinal velocity, an uncertain linear-parameter-varying (LPV) system is obtained. Based on the LPV model, a gain-scheduling observer is proposed and the observer gain can be determined with off-line computation and on-line computation. The off-line computation includes the calculation of a set of linear matrix inequalities and the on-line computation contains several algebraic operations. The proposed design methodology is applied to a four-wheel-independent-drive electric vehicle in simulation. It infers from different maneuvers that the designed observer has a good performance on estimating the sideslip angle.     

一种新型多车道车流量检测算法

为了实时有效地检测道路路口车流量信息,并为交通控制和管理提供准确的交通流数据,提出了一种新型的车流量检测算法。通过利用现有的路面标记进行图像对称分割并计算图像灰度值方差,来判断有无车辆通过,进而实现车流量计算。仿真结果表明,该算法不仅简单,易于实现,而且检测准确率高,实时性好,能够有效地为智能交通灯控制提供信息数据。     

基于局部二进制模式和图变换的快速匹配算法

针对图像匹配在图像拼接、目标识别等领域的应用中尺度不变特征变换（Scale Invariant Feature Transform,SIFT）算法计算复杂度高、实时性较差的问题,提出了一种基于局部二进制模式（Local Binary Patterns,LBP）和图变换（Graph Transformation Matching,GTM）的匹配算法.首先采用SIFT特征检测提取特征点并以特征点为中心取13×13的图像块作为特征区域;然后用本文提出的局部旋转不变二进制模式（Local Rotation Invariant Binary Patterns,LRIBP）描述子对特征区域进行描述产生29维的特征描述向量,降低了描述子的复杂度,并以欧氏距离为度量准则进行初始匹配;最后采用图变换匹配算法剔除误匹配点,从而提高算法的运算速率和匹配精度.仿真结果表明,本文所提算法不仅具有较高的精度和较强的鲁棒性,并且减少了算法的运算量,提高了算法的实时性.     

基于细胞神经网络的从阴影恢复形状的新方法

细胞神经网络(CNN)是一种实时处理信号的大规模非线性模拟电路,它的连续时间特点以及局部互连特点使其可以进行并行计算,并且非常适用于超大规模集成电路(VLSI)的实现.本文针对从阴影恢复形状(SFS)问题,提出了一种基于硬件退火CNN的能量函数优化方法,并对该方法进行了详细分析,给出了实例的仿真结果,验证了该方法的有效性.该方法为并行处理算法,具有运算量小、易于大规模VLSI集成实现,且能够克服局部极小等优点,可以使SFS问题得到实时的处理.     

Monocular precrash vehicle detection: features and classifiers.

Robust and reliable vehicle detection from images acquired by a moving vehicle (i.e., on-road vehicle detection) is an important problem with applications to driver assistance systems and autonomous, self-guided vehicles. The focus of this work is on the issues of feature extraction and classification for rear-view vehicle detection. Specifically, by treating the problem of vehicle detection as a two-class classification problem, we have investigated several different feature extraction methods such as principal component analysis, wavelets, and Gabor filters. To evaluate the extracted features, we have experimented with two popular classifiers, neural networks and support vector machines (SVMs). Based on our evaluation results, we have developed an on-board real-time monocular vehicle detection system that is capable of acquiring grey-scale images, using Ford's proprietary low-light camera, achieving an average detection rate of 10 Hz. Our vehicle detection algorithm consists of two main steps: a multiscale driven hypothesis generation step and an appearance-based hypothesis verification step. During the hypothesis generation step, image locations where vehicles might be present are extracted. This step uses multiscale techniques not only to speed up detection, but also to improve system robustness. The appearance-based hypothesis verification step verifies the hypotheses using Gabor features and SVMs. The system has been tested in Ford's concept vehicle under different traffic conditions (e.g., structured highway, complex urban streets, and varying weather conditions), illustrating good performance.     

多环境下的实时前车检测与车距测量

本文利用智能手机对视频进行实时处理,从而实现对前车的检测和车距的估算。采用两步前车检测技术,即寻找车辆假设区域和验证车辆假设区域。通过基于车道先验的两步阈值法提取车底阴影,通过车道先验对阴影进行筛选从而寻找车辆假设区域,并利用基于Haar-like特征的Adaboost方法来对车辆假设区域进行验证,之后通过视频流特征对误检进行剔除并对车辆进行跟踪。本文还结合摄像头透视几何关系,只需用户提供摄像头高度信息即可对前车距离进行准确估算,可以作为有效的危险预警手段。      

Design of a hardware accelerator for real-time moment computation:a wavefront away approach

In image processing, pattern recognition, and computer vision, one of the most powerful techniques for feature extraction is to use moments. Real-time applications of this method, however, have been prohibited due to the intensive computation encountered in calculating the moments. One solution to this problem is to adopt specially designed hardware accelerators. This paper describes, from a practical standpoint, the design of a custom hardware accelerator for speeding up the moment computation. The design of the core functional units and the design of the overall system based on a wavefront array architecture are discussed. The moment accelerator can be easily configured into different sizes to meet diverse application requirements cost effectively. Testing results based on implementation using field-programmable gate array devices show that, at an affordable cost, the proposed hardware accelerator can deliver real-time speeds for moment computation. Elimination of this computational bottleneck makes it possible to use moments-based features in real-time industrial applications     

车载自组网中可证明安全的无证书认证方案

认证协议的设计是目前车载自组网(VANET)安全领域的研究热点。现有的认证方案中普遍存在密钥托管带来的安全问题,以及使用计算量大的双线性对导致认证效率很低。针对以上问题,该文提出可证明安全的无证书批认证方案,方案中车辆的密钥由车辆自身和一个密钥生成中心共同生成,解决密钥需要托管给第三方维护的问题;方案的签名构造不使用计算量大的对运算,减少了计算开销;引入批认证来减少路边设施的认证负担,提高认证效率。基于求解椭圆曲线上的离散对数问题的困难性假设,在随机预言机模型中证明了该方案可以抵抗自适应选择消息和身份攻击,从而抵抗更改攻击和假冒攻击,并具有匿名性、可追踪性等特点。与现有方案相比,该方案实现了更高效的认证。     

Dynamic management network system of automobile detection applying edge computing

Dynamic vehicle detection requires the transmission of large amounts of data collected by different types of sensors to the edge computing nodes. This is likely to cause network delays and congestion, affecting the computation of the edge computing nodes and thus posing serious security risks. Therefore, optimizing data transmission between vehicles and edge computing nodes is a new challenge to be addressed in the practical application of edge computing-based vehicle dynamic detection architectures. The data requirements of VDT for vehicle detection dynamic detection in different environments are considered, the optimization objectives and constraints are analysed, and a deviation detection and greedy algorithm is proposed in this paper to address the problems of long mixed-integer linear programme solution time and insufficient practical applications, and the performance of the algorithm is evaluated through simulation experiments conducted by simulation of urban mobility, a traffic flow simulation tool, and PreScan, a vehicle simulation test software. The results show that compared with the deviation detection algorithm, the greedy algorithm can reduce the communication overhead by 82.6%–86.2% in all cases and improve the performance by 13.6%–19.5%, which is more suitable for practical applications. The results of this paper contribute to the automation and modernization of vehicle technology management and information transfer.     

交通监控视频中的车辆异常行为检测

构建了利用交通监控视频对车辆异常行为进行检测的系统框架.使用改进Surendra背景差分与三帧差分相结合的算法进行车辆目标检测,结合CamShift算法与Kalman滤波器进行车辆目标跟踪,提取车辆质心绘制运动轨迹,针对车辆运动方向判别、违章变道、调头等行为提出了检测方法.实验结果表明,提出的交通监控视频中的车辆异常行为检测系统具有较高的实时性与准确性,部署简易快速,维护成本低廉,可以满足当今智能交通系统日益增长的需求.     

一种适用于智能手机的图像识别算法

针对目前常用的图像识别算法运算复杂和内存占用量大,不能很好的应用于移动平台等问题,本文提出了一种适用于智能手机的图像识别算法：首先,通过使用BRISK特征点检测算法提取图像特征和低字节的FREAK描述符对特征进行表述,解决了特征检测时间长和特征描述符内存占用大的问题;其次,将智能手机的重力信息添加到图像特征中改善了BRISK特征的区分能力,解决了相似结构特征难以区分的问题;最后,建立描述符的多级索引,实现相似描述符的快速查找,解决了描述符匹配问题.实验结果表明,本文提出的算法能有效地运行在资源受限的智能手机上实现对场景的实时识别.     

基于注意力机制的实时车辆点云检测算法

针对现有激光点云目标检测效果、实时性差的问题,提出了一种基于注意力机制的实时车辆点云检测算法。本文所提出的检测算法将注意力机制算法与YOLOv3相结合,利用注意力机制对点云鸟瞰图的特征进行权重分配,以学习不同通道和空间下特征的相关性,并通过CIOU loss和Focal loss来改进检测器的损失函数。实验结果表明基于注意力机制的车辆点云检测算法检测速度可达30帧/秒,车辆目标的平均检测精度达到了92.5%。并且在实车数据测试中,该算法能快速准确的对一定范围内车辆进行准确识别,并且达到实时检测效果。     

多分辨率多模型目标跟踪方法的三处理机并行实现

介绍了一种用于噪声中机动目标跟踪的小波变换新算法——多分辨率多模型目标跟踪方法。该算法利用小波变换获取多分辨率测量数据,增大了机动检测概率,但需要几倍于传统算法的运算量,难以在实际跟踪系统中实时应用。对此,采用并行处理技术,给出了该算法在包含３个ＣＰＵ的多总线微机系统上井行实现的方法。模拟结果证明了该方法的优良性能,并行处理时间接近传统的单分辨率跟踪算法的处理时间。     

<Emphasis Type="Italic">EIAS-CP:</Emphasis> new efficient identity-based authentication scheme with conditional privacy-preserving for VANETs

In VANETs, vehicles broadcast traffic-related messages periodically according to Dedicated Short Range Communication protocol. To ensure the reliability and integrity of messages, authentication schemes are involved in VANETs. As traffic-related messages are time-sensitive, they must be verified and processed timely, or it may cause inestimable harm to the traffic system. However, the OBUs and the RSUs are limited in computation ability and cannot afford vast messages’ verification. Recently, some identity-based authentication schemes using bilinear pairing have been proposed to improve the efficiency of message verification for VANETs. Nevertheless, the bilinear pairing is not suited for VANETs due to its complex operations. The design of an efficient and secure authentication scheme with low computation cost for VANETs still is a rewarding challenge. To settle this challenge, a new efficient identity-based authentication scheme is proposed in this paper. The proposed scheme ensures reliability and integrity of messages and provides conditional privacy-preserving. Compared with the most recent proposed authentication schemes for VANETs, the computation costs of the message signing and verification in the proposed scheme reduce by 88 and 93 % respectively, while security analysis demonstrates that our proposed scheme satisfies all security and privacy requirements for VANETs.