基于卡尔曼滤波算法的城市交叉路口车辆检测及分类

为了向相关部门提供更多的过往车辆信息以满足道路交通的需求,设计了一个基于卡尔曼滤波算法的城市交叉路口车辆检测及分类系统,用于对过往的车辆进行检测、计数和分类.首先采用背景差分法和卡尔曼滤波算法对在检测区的车辆进行检测和跟踪;然后使用经过检测、处理的被测车辆图像触发距其最近的相机进行图形分割;最后,通过LDA分类器对分段车辆的几何形状及外观特征进行正确地分类.所提系统的有效性在摄取的3400帧视频序列上得到了验证,实验结果表明,系统的检测率可达97.44％,正确分类率可达88.0％,与先进的方法相比,取得了更好的检测性能.     

Perception for collision avoidance and autonomous driving

The Navlab group at Carnegie Mellon University has a long history of development of automated vehicles and intelligent systems for driver assistance. The earlier work of the group concentrated on road following, cross-country driving, and obstacle detection. The new focus is on short-range sensing, to look all around the vehicle for safe driving. The current system uses video sensing, laser rangefinders, a novel light-stripe rangefinder, software to process each sensor individually, a map-based fusion system, and a probability based predictive model. The complete system has been demonstrated on the Navlab 11 vehicle for monitoring the environment of a vehicle driving through a cluttered urban environment, detecting and tracking fixed objects, moving objects, pedestrians, curbs, and roads.     

基于车载视频图像的道路消失点检测算法

道路消失点检测是高级驾驶辅助系统中盲区监测的重要组成部分。针对现有消失点检测方法所存在的准确度低、运算量大等问题,提出一种基于车载视频图像的道路消失点检测算法。该算法在Harris角点检测基础上优化得分函数检测出图像特征点,减少在跟踪阶段的运算量;通过金字塔光流法和帧差距离对运动特征点进行跟踪,在结束帧上准确获得各特征点的位置;对特征点去除离值点后,通过优化初始聚类中心的K-Means聚类算法,得到车载视频图像的道路消失点。最后将算法应用于各种车辆行驶场景进行测试,在较短运行时间内,能准确检测出车载视频图像中道路消失点,证明算法鲁棒性好、运算简单易实现。     

New automatic vehicle identification system for detection oftraffic without lane discipline

Electronic road toll collection has usually been carried out on the highway using an automatic vehicle identification (AVI) system. At the toll site, physical barriers are set up between lanes so that individual vehicles can be effectively detected. Problems arise, however, in the detection of vehicles on busy city roads where barriers between lanes are not desirable and the traffic does not always maintain lane discipline. A new AVI system using a randomised time division multiplexing (TDM) technique for detecting several vehicles simultaneously entering the detection zone is described. Laboratory experiments carried out for up to eight identification tags at the same time showed that all replies were detected for 900 interrogations. Collisions between replying signals did not occur. The high-speed transmission of serial data has, however, an error rate of 3-7%. Under laboratory conditions, the detection errors were eliminated simply by allowing the tags to reply twice     

基于消失点迭代重估的道路边缘检测

传统道路边缘检测算法很少能用于乡村道路。该文提出基于消失点迭代重估的道路边缘检测方法:Gabor滤波器估计像素纹理方向,同时计算对应的置信度;根据纹理方向,由置信度大于阈值的像素投票确定初始消失点;以初始消失点为起点,向下建立一组射线,并计算由颜色差和二倍角正弦函数加权的方向一致率,将最大值对应的射线作为初始道路边缘;交替采样两侧边缘,选取最优消失点,由此得到两条道路边缘。实验测试了不同类型的乡村道路图片,并与经典算法对比。结果表明该方法在乡村道路中具有较高精准度与鲁棒性。     

Novel Collision Detection and Avoidance System for Midvehicle Using Offset-Based Curvilinear Motion

In recent days, the manufacture of automotive vehicles is dramatically enhanced worldwide. Most vehicle crashes are due to the drive distraction on the real highway roads and traffic-density. In this proposed method, a novel collision detection and avoidance algorithm are coined for Midvehicle Collision Detection and Avoidance System (MCDAS), addressing two scenarios, namely, (a) A rear-end collision avoidance with host vehicle under no front-end vehicle condition and (b) offset-based curvilinear motion under critical conditions, while, suitable parallel parking manoeuvring also addressed using offset-based curvilinear motion. The Monte Carlo analysis of the proposed MCDAS is demonstrated using the Constant Velocity (CV) manoeuvring strategy and simulated with real-time data using the NGSIM database.

     

Video traffic analysis using scene and vehicle models

This paper presents a complete system for analyzing a vehicle׳s behavior in the context of real-time traffic video surveillance applications. To obtain the best possible results, it is fundamental to exploit the scene characteristics and the predefined traffic rules. For that purpose, an initial training step is performed that involves estimating the geometrical structure of the road, i.e., the depth relative to the camera, the vanishing points, the road areas itself, the road decomposition (into normal and forbidden traffic lanes or areas), the traffic rules, the typical vehicle trajectories and speeds, and the lane-changing rules. This process leads to a scene model, which is used together with a simple vehicle geometrical model during the vehicle detection, tracking and trajectory estimation phases to improve the robustness against the perspective and occlusion effects. Shadow effects are also accounted for during the moving object detection phase. Finally, this spatio-temporal analysis is used to obtain information that concerns the vehicles׳ behaviors. Experiments show that the information obtained is reliable and can be computed in real time.     

一种改进的道路行车密度峰值模糊聚类算法

为解决城市道路中相邻车辆聚类精度低的问题,本文提出了一种改进的密度峰值模糊聚类算法。首先,该算法使用自适应椭圆距离代替欧式距离,并在决策图中引入指数函数曲线选择密度峰值点,以确定初始聚类中心和聚类数目;接着,将初始信息代入模糊C均值(FCM)聚类算法中,经迭代计算取得一次聚类结果;最后,根据雷达数据中同一辆车的数据点速度差值极小、不同车辆的速度差值相对较大这一特征,引入和速度相关的目标函数,并通过迭代计算取得最终的聚类结果,以对一次聚类结果进行修正。根据真实道路测量数据的实验证明,本文提出的聚类算法精度高、鲁棒性好,能正确聚类城市道路中相邻的车辆目标,具有更好的聚类效果。为道路中车辆的跟踪、交通状态预估等处理提供可靠、准确的目标信息,大大减少后续工程的计算量。     

GOLD: a parallel real-time stereo vision system for genericobstacle and lane detection

This paper describes the generic obstacle and lane detection system (GOLD), a stereo vision-based hardware and software architecture to be used on moving vehicles to increment road safety. Based on a full-custom massively parallel hardware, it allows to detect both generic obstacles (without constraints on symmetry or shape) and the lane position in a structured environment (with painted lane markings) at a rate of 10 Hz. Thanks to a geometrical transform supported by a specific hardware module, the perspective effect is removed from both left and right stereo images; the left is used to detect lane markings with a series of morphological filters, while both remapped stereo images are used for the detection of free-space in front of the vehicle. The output of the processing is displayed on both an on-board monitor and a control-panel to give visual feedbacks to the driver. The system was tested on the mobile laboratory (MOB-LAB) experimental land vehicle, which was driven for more than 3000 km along extra-urban roads and freeways at speeds up to 80 km/h, and demonstrated its robustness with respect to shadows and changing illumination conditions, different road textures, and vehicle movement.     

Propagation characteristics in open-groove waveguides surrounded by rough sidewalls

The radio propagation characteristics in groove-shaped roads surrounded by rough surface sidewalls are presented. Typical sidewalls are snowpack in heavy snow regions or buildings in urban areas. In this analysis, such roads are modeled by a straight groove waveguide with statistically rough sidewalls. A simple geometrical optical approach is applied to calculate the field strength along the longitudinal direction of the road as well as transversely across the road. The theoretical results are compared with previously presented experimental results. The theoretical results show that the surface roughness causes an additional propagation loss in regions far from the transmitting point and that the field strength in the cross section exhibits a cosine distribution with a slightly changing magnitude. These theoretical characteristics agree with the experimental results.<>     

Less is more [intelligent speed adaptation for road vehicles]

The use of sophisticated in-car technology to enforce urban speed limits and improve road safety is an idea whose time could well have come. To the great majority of motorists, speed bumps are, at best, a 'necessary evil'. However, if the results of recent trials in Sweden involving intelligent speed adaptation (ISA) technology are anything to go by then the days of this contentious feature of our urban roads could well be numbered. Between 1999 and 2002 government funding of SEK75m (#5.64m) was provided to install ISA systems in 5000 test vehicles distributed amongst four Swedish municipalities (Borlange, Lidkoping, Lund and Umea), and the results of these trials have apparently greatly exceeded the expectations of even ISA technology's keenest proponents within the Swedish National Road Administration (SNRA). ISA systems are, a combination of several existing technologies being put together in a new way. Many cars at the expensive end of the market are now being provided with satellite navigation systems as standard, where each vehicle is fitted with a Global Positioning System (GPS) transmitter/receiver, so that the precise position of the vehicle can be determined, and a GSM mobile phone link, so that this position can be integrated with a detailed and accurate digital roadmap maintained by the satellite navigation service provider. ISA technology, simply put, merely adds data on speed limits to the digital road map, so that each vehicle 'knows' what the maximum legal speed is on any road that it happens to be travelling along.     

Infrastructure‐based vehicular congestion detection scheme for V2I

With the increasing number of vehicles, traffic jam becomes one of the major problems of the fast‐growing world. Intelligent transportation system (ITS) communicates perilous warnings and information on forthcoming traffic jams to all vehicles within its coverage region. Real‐time traffic information is the prerequisite for ITS applications development. In this paper, on the basis of the vehicle‐to‐infrastructure (V2I) communication, a novel infrastructure‐based vehicular congestion detection (IVCD) scheme is proposed to support vehicular congestion detection and speed estimation. The proposed IVCD derives the safety time (time headway) between vehicles by using iterative content‐oriented communication (COC) contents. Meanwhile, the roadside sensor (RSS) provides an infrastructure framework to integrate macroscopic traffic properties into the estimation of both the traffic congestion and vehicle safety speed. The main responsibilities of RSS in IVCD are to preserve privacy, aggregate data, store information, broadcast routing table, estimate safety speed, detect traffic jam, and generate session ID (S‐ID) for vehicles. Monte Carlo simulations in four typical Chinese highway settings are presented to show the advantage of the proposed IVCD scheme over the existing Greenshield's and Greenberg's macroscopic congestion detection schemes in terms of the realized congestion detection performance. Real road traces generated by Simulation of Urban Mobility (SUMO) over NS‐3.29 are utilized to demonstrate that the proposed IVCD scheme is capable of effectively controlling congestion in both single and multilane roads in terms of density and speed health with previous schemes in this field.     

基于光学多传感器场景信息的视觉辅助驾驶技术

为满足坦克、装甲车辆等军用车辆的闭舱、无窗驾驶需求,研制了一套新型辅助驾驶系统。系统将分布于车辆四周的多路光学传感器获取车辆近身场景,通过全景拼接算法得到车辆近身360°的全景鸟瞰视频,该视频显示于车载显示屏,用于车辆通过窄道、有障碍物等特殊路段,或倒车时,驾驶员观看。同时,在车辆通过常规路段时,上述视频可根据驾驶员头部扭转角度,裁选出符合人眼观察视角的车外场景视频,传输至驾驶员显示头盔上,供驾驶员观看。如遇特殊情况,车载显示屏会报警,驾驶员将回归车载显示屏的观看。其中,驾驶员头部位置确定方法采用了红外LED光源图像定位技术和MEMS惯性器件定位技术。在实验室,搭建模型小车,验证全景鸟瞰视频生成技术和头盔自由视点观察技术。此外,还使用真实车辆进行了跑车实验。实验结果表明,上述系统可满足闭舱无窗车辆在常规路况下行驶,速度可达40 km/h,同时可辅助车辆窄道行车、障碍物绕行和倒车等事项顺利进行。     

PLC在交通信号控制系统中的应用

随着我国城市化和现代化建设的迅速发展,交通问题成为社会经济发展的一个大问题。原有的信号灯定时控制方法已经无法满足越来越拥堵道路的需求。城市道路交通信号PLC智能控制系统运行可靠、操作简便、应用范围广,能够有效降低车辆延误及停车的次数,缓解道路交通拥堵情况。本文基于S7-200系列PLC控制系统,根据时间指令将24小时分为:白天控制与夜间控制两个时间段,并运用了实时检测、判断延时,实现交通灯信号的智能控制,有效缓解交通高峰期的压力和主干道路口的拥挤现状,推动我国城市道路交通信号灯控制系统的科学化、智能化发展。     

Comparison between braking and steering yaw moment controllers considering ABS control aspects

Two yaw motion control systems that improve a vehicle lateral stability are proposed in this study: a braking yaw motion controller (BYMC) and a steering yaw motion controller (SYMC). A BYMC controls the braking pressure of the rear inner wheel, while a SYMC steers the rear wheels to allow the yaw rate to track the reference yaw rate. A 15 degree-of-freedom vehicle model, simplified steering system model, and driver model are used to evaluate the proposed BYMC and SYMC. A robust anti-lock braking system (ABS) controller is also designed and developed. The performance of the BYMC and SYMC are evaluated under various road conditions and driving inputs. They reduce the slip angle when braking and steering inputs are applied simultaneously, thereby increasing the controllability and stability of the vehicle on slippery roads. The SYMC performs better than the BYMC because the SYMC vehicle has four-wheel steering. However, both the BYMC and SYMC vehicles show improved performance during lane-change maneuvers.     

Automatic Moving Vehicle Detection and Classification Based on Artificial Neural Fuzzy Inference System

The main objective of this work is to automatically detect moving vehicles on the road. Adaptive Neuro-Fuzzy Inference System (ANFIS) classifier is adopted in this paper to classify moving vehicles on road. An input traffic video scenes are taken with vertical and horizontal positioned cameras. The proposed system contains six major steps such as preprocessing, vehicle detection, tracking, structural matching, feature extraction and classification. In this proposed method, preprocessing consists of color conversion and noise removal. Vehicle detection is performed by using background subtraction and Otsu thresholding algorithm. Kalman filter is used in the third step to track moving vehicles in successive frames. In the fourth step, Active Shape Modelling method is used to recover the 3D shape of the vehicle in order to find the boundaries of vehicle. In the fifth step, features of the detected vehicles are extracted by Harrish corner detector, log Gabor filter and these features are taken into account to classify the types of vehicle. Finally, ANFIS is proposed to classify the vehicles which is trained by updating the membership function. Experimentation results provides better accuracy rate and low mean error rate when compared with the state of art methods.     

Performance Analysis with Traffic Accident for Cooperative Active Safety Driving in VANET/ITS

Recently, by using vehicle-to-vehicle and vehicle-to-infrastructure communications for VANET/ITS, the cooperative active safety driving (ASD) providing vehicular traffic information sharing among vehicles significantly prevents accidents. Clearly, the performance analysis of ASD becomes difficult because of high vehicle mobility, diverse road topologies, and high wireless interference. An inaccurate analysis of packet connectivity probability significantly affects and degrades the VANET/ITS performance. Especially, most of related studies seldom concern the impact factors of vehicular accidents for the performance analyses of VANET/ITS. Thus, this paper proposes a two-phase approach to model a distributed VANET/ITS network with considering accidents happening on roads and to analyze the connectivity probability. Phase 1 proposes a reliable packet routing and then analyzes an analytical model of packet connectivity. Moreover, the analysis is extended to the cases with and without exhibiting transportation accidents. In phase 2, by applying the analysis results of phase 1 to phase 2, an adaptive vehicle routing, namely adaptive vehicle routing (AVR), is proposed for accomplishing dynamic vehicular navigation, in which the cost of a road link is defined in terms of several critical factors: traffic density, vehicle velocity, road class, etc. Finally, the path with the least path cost is selected as the optimal vehicle routing path. Numerical results demonstrate that the analytical packet connectivity probability and packet delay are close to that of simulations. The yielded supreme features justify the analytical model. In evaluations, the proposed approach outperforms the compared approaches in packet connectivity probability, average travel time, average exhausted gasoline. However, the proposed approach may lead to a longer travel distance because it enables the navigated vehicle to avoid traversing via the roads with a higher traffic density.     

Extended MULE Concept for Traffic Congestion Monitoring

Wireless Sensor Networks are being recently studied to monitor real-time traffic conditions on roads and highways. Idea of using vehicles to convey information from sensors placed alongside roads to the dedicated base stations has also been under scrutiny for some time. In this paper, we argue that a sensor placed on a vehicle instead of a fixed location can effectively sense traffic congestion on the road and report it to the already available WLAN Access Points (APs) instead of the dedicated base stations. This way, instead of deploying series of base stations to collect traffic information, congestion information can be sent over the ISM links between the vehicular sensor nodes and the WLAN APs. This paper investigates, as we call it, the Extended MULE concept by using actual experimental data obtained from the test drives across the city. Our results show that adopting this idea is effective in reporting traffic congestion on the roads.     

基于三维GIS的城市高架环线噪声地图研究

随着中国城市道路交通的迅速发展,城市环线交通以高架环线为主。现有的各类道路噪声预测模型都缺乏处理城市高架环线这类复杂的空间三维实体的能力。以美国联邦公路局的FHWA噪声预测模型为基础,通过车辆等效计算、高架环线多层多车道修正、高架桥声屏障修正等措施,对FHWA噪声预测模型进行修正。并将修正后的噪声预测模型与三维GIS模型相结合,绘制出了城市高架环线噪声地图。最终通过与实测数据的对比,验证了所修正噪声模型的正确性。     

基于道路的运动车辆检测

基于图像的运动车辆检测是智能交通系统中一个重要而又困难的问题，它通常受到光照、路边摇晃树木、以及各种恶劣天气等的影响。提出了一种基于道路的车辆检测方法，在道路的入口处只利用背景差，对非入口处运动背景和不同道路上运动车辆的运动向量分别进行高斯建模，并利用贝叶斯准则和多帧信息来检测出运动车辆和运动背景。实验结果显示该方法取得了很好的效果。