By Topic

Intelligent Vehicles Symposium, 2008 IEEE

Date 4-6 June 2008

Filter Results

Displaying Results 1 - 25 of 201
  • Real-time drowsiness detection system for an intelligent vehicle

    Page(s): 637 - 642
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (684 KB) |  | HTML iconHTML  

    In the last years, the traffic accidents study is become important because they produce several died and hurt around the world. To help in reducing this fatality, in this paper, a new advanced driver assistance system (ADAS) for automatic driver's drowsiness detection based on visual information and artificial intelligent is presented. This system works on several stages to be fully automatic. In addition, the aim of this algorithm is to locate and to track the face and the eyes to compute a drowsiness index. Examples of different driver's images taken over real vehicle are shown to validate the algorithm that works in real time. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • The ActMAP - FeedMAP framework A basis for in-vehicle ADAS application improvement

    Page(s): 263 - 268
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (424 KB) |  | HTML iconHTML  

    Up to date digital maps are a demanding requirement especially in the context of digital map based ADAS applications. This paper presents first results and applications from the FeedMAP project and how they can be used for increasing driving safety by integrating map deviation detection and incremental update technology into ADAS frameworks using the ADAS horizon concept. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Vehicle/pedestrian conflict analysis and exclusive right-turn phase setting study

    Page(s): 733 - 738
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (775 KB) |  | HTML iconHTML  

    Many Asian cities, i.e. Beijing, China, are facing increasing conflicts between the booming vehicles and bicyclists/pedestrians in the five years. One important alleviating method is to introduce exclusive right-turn phases which have rarely been designed or set up for signalized intersections before. To estimate how this method can improve traffic efficiency, the conflicts between right-turn vehicles and other road users are first studied and modeled through vision monitoring. Then, different settings of phasing scheme are compared based on the simulation tests that are designed to incorporate the conflicts characteristics. The results show that the vehicle/pedestrian traffic efficiency varies with several conditions, i.e. vehicle/pedestrian arriving rate. This indicates that some appropriate adaptive phasing strategies should be used to better serve traffic requirements. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Majority-game based conflict modeling for pedestrians and right-turning vehicles in signalized intersection

    Page(s): 1191 - 1196
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (314 KB) |  | HTML iconHTML  

    Interactions between the crossing pedestrians and right-turning, left-turning, or through vehicles are frequently seen in cities. The resulting crowd dynamics is usually complex and of great difficulty to pedestrian clearance time estimation and signal phase optimization. Based on the study of abundant video data of competitive crossing, an elaborate but understandable cellular automata model is proposed in this paper to describe the competition behaviors of the pedestrians and right-turning drivers, when they are trying to pass the same crosswalk. The interaction between the drivers and pedestrians are viewed and simulated as a majority game between drivers and pedestrians. The simulation results show that this computationally efficient microscopic model fits well with the observed delay of vehicles and pedestrians. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Experimental evaluation of a novel vehicular communication paradigm based on cellular networks

    Page(s): 198 - 203
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (809 KB) |  | HTML iconHTML  

    In the field of vehicular networks, the amount of telematic services which are usually taken into account is very limited. Safety services, and specifically collision avoidance applications, practically receive an exclusive attention. Due to vehicular ad-hoc networks (VANETs) are the most used communication technology, services conceived for the vehicle domain are frequently designed to take advantage of its benefits, but also to suffer its limitations. The intention of this paper is proposing a novel communication paradigm open to the development of any vehicular service with connectivity requirements. This way, not only vehicle to vehicle (V2V) necessities are considered, but also vehicle to infrastructure (V2I) connections are taken into account with the same importance. The work presented here chooses the cellular networks as a valid alternative to VANET approaches in most of the cases, with the added value of V2I capabilities. A design based on peer to peer (P2P) networks has been implemented and tested over a real environment. The hardware/software prototype is explained and main performance measurements prove our system is a feasible communication paradigm for most of vehicular services. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Analysis of driver behavior based on traffic incidents for driver monitor systems

    Page(s): 930 - 935
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (482 KB) |  | HTML iconHTML  

    Research is being conducted into a large number of on-board driver monitor systems as a means of reducing traffic accidents. In order to improve the effectiveness of these systems, it is necessary to detect the driver behavior and mental and physical state immediately before an accident, and to inform or warn the driver of the danger, or else to send an intervention signal to the pre-crash safety system and other advanced vehicle safety systems. Previous research has been conducted for conditions of apparent risk, and has used drive recorders to analyze the causes of accidents and to investigate and analyze driver behavior and other factors which are present immediately before an accident. This study involved an investigation of near-miss accidents (hereafter referred to as ldquoincidentsrdquo) by means of interviews in order to determine the driver behavior and mental and physical state immediately before the incident, when there was the potential risk of an accident. The purpose of this study is to contribute to research concerning advanced vehicle safety systems. We will here provide an analysis of the results and propose direction for future research concerning driver monitor systems. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Stochastic reachable sets of interacting traffic participants

    Page(s): 1086 - 1092
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (328 KB) |  | HTML iconHTML  

    Knowledge about the future development of a certain road traffic situation is indispensable for safe path planning of autonomous ground vehicles or action selection of intelligent driver assistance systems. Due to a significant uncertainty about the future behavior of traffic participants, the prediction of traffic situations should be computed in a probabilistic setting. Under consideration of the dynamics of traffic participants, their future position is computed probabilistically by Markov chains that are obtained with methods known from hybrid verification. The characteristic feature of the presented approach is that all possible behaviors of traffic participants are considered, allowing to identify any dangerous future situation. The novel contribution of this work is the explicit modeling of the interaction of traffic participants, which leads to a more accurate prediction of their positions. Results are demonstrated for different traffic situations. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Traffic sign classification using invariant features and Support Vector Machines

    Page(s): 530 - 535
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (431 KB) |  | HTML iconHTML  

    This paper presents a novel approach to recognize traffic signs using invariant features and support vector machines (SVM). Images of traffic signs are collected by a digital camera mounted in a vehicle. They are color segmented and all objects which represent signs are extracted and normalized to 36 x 36 pixels images. Invariant features of sign rims and speed-limit sign interiors of 350 and 250 images are computed and the SVM classifier is trained with these features. Two stages of SVM are trained; the first stage determines the shape of sign rim and the second determines the pictogram of the sign. Training and testing of both SVM classifiers are done using still images. The best performance achieved is 98% for sign rims and 93% for speed limit signs. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A computer vision-based system for real-time detection of sleep onset in fatigued drivers

    Page(s): 25 - 30
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (429 KB) |  | HTML iconHTML  

    This paper proposes a novel approach for the real-time detection of sleep onset in fatigued drivers. Sleep onset is the most critical consequence of fatigued driving, as shown by statistics of fatigue-related crashes. Therefore, unlike previous related work, we separate the issue of sleep onset from the global analysis of the physiological state of fatigue. This allows us for formulating our approach as an event-detection problem. Real-time performance is achieved by focusing on a single visual cue (i.e. eye-state), and by a custom-designed template-matching algorithm for on-line eye-state detection. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Towards a hardware-based system for real-time vehicle tracking

    Page(s): 385 - 390
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (198 KB) |  | HTML iconHTML  

    Real-time vehicle tracking in ITS (intelligent transportation systems) is highly desirable. In this paper, we propose a hardware-based tracking system to achieve real-time vehicle tracking. The proposed tracking system employs two innovative designs. One is that a tracking algorithm based on motion estimation is adopted. Motion estimation is achieved by full-search block matching algorithm. The other innovation is that motion estimation is implemented in hardware, which greatly reduces computational time compared to the traditional software approach. Once the motion vectors are obtained, the remaining tracking process is implemented in software. Experiments have validated the tracking algorithm and simulations have shown that the proposed tracking system can achieve real-time tracking up to 40 frames per second. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Advanced transmission cycle control scheme for autonomous decentralized TDMA protocol in safe driving support systems

    Page(s): 1062 - 1067
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (1576 KB) |  | HTML iconHTML  

    When the volume of traffic on a roadway is so high such as traffic jam, medium access control (MAC) protocols in inter-vehicle communication (IVC) systems do not work well. The primary reason is the huge amount of communication traffic, which could be resolved by conventional transmission cycle control (TCC) schemes. However, when using TCC schemes, the failure of packet transmission extends the communication interval between vehicles, which can cause driving support systems that use IVC to malfunction. This paper proposes an advanced TCC scheme for realizing a safe driving support system. The proposed scheme exploits the frame-information exchange in the autonomous decentralized TDMA protocol, so that subsequent packets are transmitted rapidly even when failure occurs. Such a scheme is shown to be effective in terms of both communication quality and driving support systems. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Can subclasses help a multiclass learning problem?

    Page(s): 214 - 219
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (895 KB) |  | HTML iconHTML  

    There are many issues that are extensively studied in a multiclass learning problem, e.g. classifier selection, data balancing, training schemes, etc. In this paper, we introduce the subclass partition and present it as a novel factor that will influence the multiclass learning performance. In many multiclass learning problems, the implicit subclass subsumption information is often ignored. This paper studies the role of the subclass and outlines the connection between the discrimination boundary and the subclass partition of the classifier. The paper investigates the use of the subclass partition to help design the better classifier architecture and formalizes the subclass partition by introducing the partition space and the partition search tree. It also proposes an algorithm to heuristically search for the optimal subclass partition. We apply the proposed scheme to a 3-class occupant classification problem, which includes 16 subclasses. The experiments are positive to demonstrate that we can improve the multiclass learning performance by searching for the optimal output class partitions. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Collision sensing by stereo vision and radar sensor fusion

    Page(s): 404 - 409
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (566 KB) |  | HTML iconHTML  

    To take the advantages of both stereo cameras and radar, this paper proposes a fusion approach to accurately estimate the location, size, pose and motion information of a threat vehicle with respect to the host from observations obtained by both sensors. To do that, we first fit the contour of a threat vehicle from stereo depth information, and find the closest point on the contour from the vision sensor. Then the fused closest point is obtained by fusing radar observations and the vision closest point. Next by translating the fitted contour to the fused closest point, the fused contour is obtained. Finally the fused contour is tracked by using the rigid body constraints to estimate the location, size, pose and motion of the threat vehicle. Experimental results from both the synthetic data and the real world road test data demonstrate the success of the proposed algorithm. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Eye gaze and movement behaviour in the operation of adaptive in-car touchscreens

    Page(s): 1027 - 1032
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (971 KB) |  | HTML iconHTML  

    The usability of in-car touchscreens with small display areas can be significantly improved if buttons are only displayed or enlarged once the userpsilas hand approaches the display. This survey examines the influence of such adaptive buttons on eye and hand movements in the vehicle. The results show that movement behaviour in the use of both static and adaptive interfaces comprises a number of movement phases whose interplay can be described with the Optimized Initial Impulse Model (OIIM). The adaptive expansion of buttons leads to significantly lower error rates compared with non-adaptive buttons, given comparable movement times. A particular observation was the influence of indicators (visible information on the relevant target prior to start of movement) during the use of adaptive interfaces. It emerges that these contribute to a reduction of the movement times, as they make motion planning easier and thereby significantly increase the precision of the initial movement phase. In addition, this means a clear reduction in eye gaze duration. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Vehicle detection by edge-based candidate generation and appearance-based classification

    Page(s): 428 - 433
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (723 KB) |  | HTML iconHTML  

    This paper presents a monocular machine vision system capable of detecting vehicles in front or behind of our own vehicle. The system consists of two main steps: 1) generation of candidates with respect to a vehicle by analyzing textures, 2) verification of the candidates by an appearance-based method using the AdaBoost learning algorithm. The vehicle candidates are generated by exploiting the facts that a vehicle has vertical and horizontal lines, and furthermore the rear and frontal shapes of a vehicle show symmetry. The proposed system is proven to be effective through experiments under various traffic scenarios. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Implementation considerations for single-camera steering assistance systems on a fixed point DSP

    Page(s): 697 - 702
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (237 KB) |  | HTML iconHTML  

    The design flow of camera-based steering assistance algorithms usually begins with their implementation in floating-point on a PC or workstation. This abstraction from all implementation effects allows an exploration of the algorithm space. Memory, throughput and word-length requirements may not be important issues for offline implementation of the algorithms, but they can become critical issues for real-time implementations on embedded processors. The implementation of driver assistance systems is faced with practical constraints because these algorithms usually need to run in real-time on fixed point digital signal processors (DSPs) to reduce total hardware cost. In this paper we first evaluate numerical requirements for implementation of camera-based lateral position detection algorithms, such as lane keep assistant and lane departure warning. We then present methods that address the challenges and requirements of fixed-point design process. The flow proposed is targeted at converting C/C++ code with floating-point operations into C code with integer operations that can then be fed through the native C compiler for a fixedpoint DSP. We demonstrate the flow on tracking example (extended Kalman filter) using synthetically generated data, and we analyze trade-offs for algorithm implementation in fixed-point arithmetic. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Multi-modal real-world driving data collection, transcription, and integration using Bayesian Network

    Page(s): 150 - 155
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (387 KB) |  | HTML iconHTML  

    In this paper we present our on-going data collection of multi-modal real-world driving. Video, speech, driving behavior, and physiological signals from 150 drivers have already been collected. To provide a more meaningful description of the collected data, we propose a transcription protocol based on six major groups: driver mental state, driver actions, driverpsilas secondary task, driving environment, vehicle status, and speech/background noise. Data from 30 drivers are transcribed. We then show how transcription reliability can be improved by properly training annotators. Finally, we integrate transcriptions, driving behavior, and physiological signals using a Bayesian network for estimating a driverpsilas level of irritation. Estimations are compared to actual values, assessed by the drivers themselves. Preliminary results are very encouraging. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Night time vehicle detection for driving assistance lightbeam controller

    Page(s): 291 - 296
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (318 KB) |  | HTML iconHTML  

    In this paper we present an effective system for detecting vehicles in front of a camera-assisted vehicle (preceding vehicles traveling in the same direction and oncoming vehicles traveling in the opposite direction) during night time driving conditions in order to automatically change vehicle head lights between low beams and high beams avoiding glares for the drivers. Accordingly, high beams output will be selected when no other traffic is present and will be turned on low beams when other vehicles are detected. Our systemuses a B&W micro-camera mounted in the windshield area and looking at forward of the vehicle. Digital image processing techniques are applied to analyze light sources and to detect vehicles in the images. The algorithm is efficient and able to run in real-time. Some experimental results and conclusions are presented. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Testbed for wireless vehicle communication: a simulation approach based on three-phase traffic theory

    Page(s): 180 - 185
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (228 KB) |  | HTML iconHTML  

    A testbed for wireless vehicle communication based on a microscopic model in the framework of three-phase traffic theory is presented. In this testbed, vehicle motion in traffic flow and analyses of a vehicle communication channel access based on IEEE 802.11e mechanisms, radio propagation modeling, message reception characteristics as well as all other effects associated with ad-hoc networks are integrated into a three-phase traffic flow model. Thus simulations of both vehicle ad-hoc network and traffic flow are integrated onto a single testbed and perform simultaneously. This allows us to make simulations of ad-hoc network performance as well as diverse scenarios of the effect of wireless vehicle communications on traffic flow during simulation times, which can be comparable with real characteristic times in traffic flow. In addition, the testbed allows us to simulate cooperative vehicle motion together with various traffic phenomena, like traffic breakdown at bottlenecks. Based on simulations of this testbed, some statistical features of ad-hoc vehicle networks as well as the effect of C2C communication on increase in the efficiency and safety of traffic are studied. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Implementation of road traffic signs detection based on saliency map model

    Page(s): 542 - 547
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (1136 KB) |  | HTML iconHTML  

    In this paper, we proposed a new road traffic sign detection model based on human-like selective attention mechanism for implementing interactive workload manager system. Since the road traffic sign boards have dominant color contrast against backgrounds, we consider the color opponents and its edge information with center surround difference and normalization as a pre-processing, which is effective to intensify the sign board color characteristics as well as reduce background noise influence. After constructing the road traffic sign saliency map using the edge and color feature maps, the candidate road traffic sign regions are selected by local maximum energy searching with entropy maximization algorithm to find suitable size of the sign board areas. Computational experiment results show that the proposed model can successfully detect a road traffic sign board. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Probabilistic approach for modeling and identifying driving situations

    Page(s): 343 - 348
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (730 KB) |  | HTML iconHTML  

    Intelligent vehicles need increasing knowledge about both their own state and the driving environment. In this work a novel method for interpreting this information by a reliable detection of relevant driving situations and driving maneuvers is proposed. The information of a situation or maneuver is extracted and hence provided for subsequent processing in the applications. As a result of different situation perception and maneuver realization of the drivers, the selected method is based on probabilistic decisions. Furthermore the inaccuracy of this decision is estimated by the inaccuracies of the sensor measurements. This value can be seen as quality measure of the probabilistic situation and maneuver detection. In addition the model allows to derivate requirements on the sensors, while determining a relevance ranking of the separate sensor information regarding the situation decision. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Color vision-based multi-level analysis and fusion for road area detection

    Page(s): 602 - 607
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (1072 KB) |  | HTML iconHTML  

    This paper presents an integrated approach to robust analysis of road area images in front of the car from a single color camera. In order to get more information from the image source, we build a three-level data fusion based on Dempster-Shaferpsilas decision theory. During the first level, we separate the input image into perspective view and birdpsilas view and recognize small road patches from the background using color histogram features. In the second level we fuse the evidences from each recognized patch and re-calculate the road-like probabilities in each view. The third level is to merge the results of the two views into one birdpsilas-view segmentation map to achieve a high degree of reliability. Our Bagging-like method has been tested in real world environments and the fusion results have been shown to be robust to road textures, trees and moving objects in the scene. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Real time pedestrian detection by fusing PMD and CMOS cameras

    Page(s): 925 - 929
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (760 KB) |  | HTML iconHTML  

    In this work we present the preliminary results of the fusion of photonic mixer device - PMD and CMOS cameras for driver assistance applications. Although the algorithms are demonstrated mainly for pedestrians, they apply to the other objects on the street. PMD camera delivers the 3D object list. Object coordinates are further projected into CMOS image plane where classification is performed using support vector machines. As compared to PMD camera the CMOS camera has higher resolution, which gives the possibility to realize finer object detection, separation and classification. As the feature vector we use quadruple haar discrete wavelet transformation (QH DWT). The speed improvement of the SVM in the testing phase (necessary for real-time implementation) is realized with Burgpsilas reduced set vector method (BRSVM), improving classification speed nearly 70 times. We have achieved the pedestrian detection rate of 80%. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A two-staged approach to vision-based pedestrian recognition using Haar and HOG features

    Page(s): 554 - 559
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (367 KB) |  | HTML iconHTML  

    This article presents a two-staged approach to recognize pedestrians in video sequences on board of a moving vehicle. The system combines the advantages of two feature families by splitting the recognition process into two stages: In the first stage, a fast search mechanism based on simple features is applied to detect interesting regions. The second stage uses a computationally more expensive, but also more accurate set of features on these regions to classify them into pedestrian and non-pedestrian. We compared various feature extraction configurations of different complexities regarding classification performance and speed. The complete system was evaluated on a number of labeled test videos taken from real-world drives and also compared against a publicly available pedestrian detector. This first system version analyzes only single image frames without using any temporal information like tracking. Still, it achieves good recognition performance at reasonable run time. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Using utility and microutility for information dissemination in Vehicle Ad Hoc Networks

    Page(s): 755 - 762
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (287 KB) |  | HTML iconHTML  

    We describe an approach to propagating streams of information in Vehicle Ad-Hoc Networks (VANETs) based on sources of information anticipating where their information will be useful. In this paper we describe how sources can model the potential usefulness of their information using utility functions. These utility functions are converted to more compact ldquomicroutilitiesrdquo that travel with the individual data packets. The microutilities allow the information forwarding protocols to operate distributedly and independently on individual data packets, while achieving good overall coordination and delivery for entire data streams. We describe the algorithms used to convert utility functions to microutilities. Our algorithms insure that both proactive planning and reactive dropping of information in-transit are done consistent with the needs of the different applications. In this way data streams from both high priority (safety) and lower priority (traffic and commercial) applications can be propagated in the same network. We show experimental results that demonstrate the advantage of such a utility.microutility approach in serving the needs of diverse intelligent transportation system applications. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.