Development of a Vision-Based Driver Assistance System with Lane Departure Warning and Forward Collision Warning Functions

Because of the increasing number of population of vehicles, the road traffic accidents are becoming more and more serious in recent years. Hence, improving driver assistance systems for security has become an important area of research. This paper presents a robust lane detection and tracking system based on monocular vision. We use the Lane Departure Warning (LDW) systems to detect the position of the vehicle with respect to the lane boundary. An algorithm to detect the road lane marking and to control the direction of the vehicle is proposed. Our work consists in establishing a Region Of Interest (ROI) of the road images, data pre-processing using the Gaussian filter and then apply the Canny edge detector to enhance lane boundaries. A method to extract lane boundaries based on color information and image segmentation by using the histogram threshold, Hough transform is proposed, and the current vehicle position is obtained. In such a case, we can decide if the vehicle is doing lane departure used on the vehicle's position, the system sends a warning message to the driver. Our proposed algorithm works accurately with various lighting conditions as well as on different road types.

A technology for unintended lane departure warning was proposed. As crucial information, lane boundaries were detected based on principal component analysis of grayscale distribution in search bars of given number and then each search bar was tracked using Kalman filter between frames. The lane detection performance was evaluated and demonstrated in ways of receiver operating characteristic, dice similarity coefficient and real-time performance. For lane departure detection, a lane departure risk evaluation model based on lasting time and frequency was effectively executed on the ARM-based platform. Experimental results indicate that the algorithm generates satisfactory lane detection results under different traffic and lighting conditions, and the proposed warning mechanism sends effective warning signals, avoiding most false warning.

Objective: Single-vehicle collisions involve only 10 percent of all occupants in crashes in the United States, yet these same crashes account for 31 percent of all fatalities. Along with other vehicle safety advancements, lane departure warning (LDW) systems are being introduced to mitigate the harmful effects of single-vehicle collisions. The objective of this study is to quantify the number of crashes and seriously injured drivers that could have been prevented in the United States in 2012 had all vehicles been equipped with LDW.Methods: In order to estimate the potential injury reduction benefits of LDW in the vehicle fleet, a comprehensive crash and injury simulation model was developed. The model's basis was 481 single-vehicle collisions extracted from the NASS-CDS for year 2012. Each crash was simulated in 2 conditions: (1) as it occurred and (2) as if the driver had an LDW system. By comparing the simulated vehicle's off-road trajectory before and after LDW, the reduction in the probability of a crash was determined. The probability of a seriously injured occupant (Maximum Abbreviated Injury Score [MAIS] 3+) given a crash was computed using injury risk curves with departure velocity and seat belt use as predictors. Each crash was simulated between 18 and 216 times to account for variable driver reaction, road, and vehicle conditions. Finally, the probability of a crash and seriously injured driver was summed over all simulations to determine the benefit of LDW.Results and Conclusions: A majority of roads where departure crashes occurred had 2 lanes and were undivided. As a result, 58 percent of crashes had no shoulder. LDW will not be as effective on roads with no shoulder as on roads with large shoulders. LDW could potentially prevent 28.9 percent of all road departure crashes caused by the driver drifting out of his or her lane, resulting in a 24.3 percent reduction in the number of seriously injured drivers.The results of this study show that LDW, if widely adopted, could significantly mitigate a harmful crash type. Larger shoulder width and the presence of lane markings, determined by manual examination of scene photographs, increased the effectiveness of LDW. This result suggests that highway systems should be modified to maximize LDW effectiveness by expanding shoulders and regularly painting lane lines.

Advanced Driver Assistance Systems (ADAS) have emerged as a critical technology in enhancing vehicle safety, providing convenience to drivers, and facilitating the transition toward fully autonomous vehicles. The core of these systems relies heavily on computer vision (CV) technologies, which enable vehicles to interpret and understand their surrounding environment in real time. The integration of computer vision in ADAS allows vehicles to recognize and process various objects such as pedestrians, vehicles, traffic signs, and lane markings, significantly improving both safety and the driving experience. In this paper, we explore the pivotal role of computer vision in the development and implementation of ADAS in autonomous vehicles. We begin by reviewing the evolution of ADAS, detailing the different sub-systems such as lane departure warning, automatic emergency braking, adaptive cruise control, and parking assistance, all of which depend on sophisticated image processing techniques. Furthermore, the paper examines the intersection of computer vision with other emerging technologies such as machine learning, deep learning, and sensor fusion. We discuss how these innovations enable ADAS to function more effectively in complex, dynamic environments. The methodology section focuses on cutting-edge algorithms used for object detection, semantic segmentation, and real-time image processing, which are integral to the functionality of modern ADAS. Additionally, we evaluate the challenges that computer vision systems face in real-world applications, including issues with sensor calibration, environmental conditions, computational limitations, and the integration of data from multiple sensors (e.g., cameras, LiDAR, radar). The paper also presents an analysis of the results obtained from various real-world implementations of ADAS technologies, highlighting improvements in driver safety and the reduction of traffic-related accidents. We explore the potential of computer vision in enabling fully autonomous driving by addressing the current gaps in technology and the advancements needed to close these gaps. Finally, the paper concludes with a discussion on the future of ADAS and the role of computer vision in realizing fully autonomous, reliable, and efficient transportation systems. We argue that continued research and development in computer vision, combined with innovations in deep learning and sensor technologies, are critical to advancing the capabilities and reliability of autonomous vehicles in the near future.

In this paper, we present a deep neural network based real-time integrated framework to detect objects, lane markings, and drivable space using a monocular camera for advanced driver assistance systems. The object detection framework detects and tracks objects on the road such as cars, trucks, pedestrians, bicycles, motorcycles, and traffic signs. The lane detection framework identifies the different lane markings on the road and also distinguishes between the ego lane and adjacent lane boundaries. The free space detection framework estimates the drivable space in front of the vehicle. In our integrated framework, we propose a pipeline combining the three deep neural networks into a single framework, for object detection, lane detection, and free space detection simultaneously. The integrated framework is implemented in C++ and runs real-time on the Nvidia's Drive PX 2 platform.

ABSTRACTObjective: There are little objective data on whether drivers with lane departure warning and forward collision warning systems actually use them, but self-report data indicate that lane departure warning may be used less and viewed less favorably than forward collision warning. The current study assessed whether the systems were turned on when drivers brought their vehicles to dealership service stations and whether the observational protocol is a feasible method for collecting similar data on various manufacturers' systems.Methods: Observations of 2013–2015 Honda Accords, 2014–2015 Odysseys, and 2015 CR-Vs occurred at 2 U.S. Honda dealerships for approximately 4 weeks during Summer 2015.Results: Of the 265 vehicles observed to have the 2 systems, 87 (32.8%) had lane departure warning turned on. Accords were associated with a 66% increase in the likelihood that lane departure warning was turned on compared with Odysseys, but the rate was still only about 40% in Accords. In contrast, forward collision warning was turned on in all but one of the observed vehicles.Conclusions: Observations found that the activation rate was much higher for forward collision warning than lane departure warning. The observation method worked well and appears feasible for extending to other manufacturers.

This paper focuses on a night time driving performance evaluation of an image-based lane departure warning (LDW) system which was previously developed by the author [25, 26]. In the previous study, road testing was only conducted under the day time driving condition. So, the main focus here is to conduct night time driving evaluations. This system uses the Lucas-Kanada (L-K) optical flow and the Hough transform methods approach. Based on the status of the vehicle deviating from its heading lane, the method integrates both techniques to establish an operation algorithm to determine whether a warning signal should be issued. The L-K optical flow tracking is used when the lane boundaries cannot be detected, while the lane detection technique is used when they become available. Even though both techniques are used in the system, only one method is activated at any given time because each technique has its own advantages and also disadvantages. The image-based LDW system was road tested on rural highways and this paper briefly presents our night time driving test results.

Using multisensory signals in advanced driver assistance is continuously increasing as a way to increase the attention and reduce the reaction time. It is essential that driver assistance system is capable of providing directional cues to the driver to direct his attention to the sides of the car as usually the focus is in front of the car. These signals could be for blind spot information, navigation, lane departure warning, collision warning, etc. This study investigated the effect of auditory and vibrotactile on directional attention in driver assistance systems. Moreover, two types of immersive displays were used in the driving simulation, namely the Head Mounted Display (HMD) and CAVE display, to study the effect of the type of display on the human performance. Lane Chang Task was used to assess the attention by measuring the response time to directional information. Vibrotactile and Auditory cues induced equal response times, meanwhile, vibrotactile signal was significantly gained higher satisfaction than auditory.

Advanced Driver Assistance System (ADAS) is a very important part of an up to date vehicle. For achieving highlevel objectives in such ADAS functionality like LKA (lane keeping assistance), LDW (lane departure warning system), FCW (forward collision warning) the quality of the algorithms under the hood must be extremely high. In the last few years, it is common that these algorithms are based on DNNs (deep neural networks) applied to the tasks of semantic and instance segmentation, 2D/3D object detection and visual object tracking. Recent state-of-the-art DNN models as usual solve only one single task from the listed above and running several neural networks is rather computationally expensive and even impossible due to the lack of the GPU memory. One of the approaches used to overcome such a problem is a shared backbone (also called feature extractor or encoder). The backbone consumes most of the computing resources thus the model with a shared backbone achieves better inference performance. Unfortunately, the training procedure for a shared backbone model has several difficulties. The first one is the lack of datasets with all the required and uniform annotation types. The second problem is a more sophisticated backpropagation procedure. In this paper, we consider several methods for multi-task neural network training and present the results of such training procedures on several public datasets with dissimilar annotation types. The shared backbone is applied to the following three tasks performed simultaneously on the road scene: semantic segmentation, 2D object detection and 3D object detection. While the performance of the DNNs with shared backbone increased significantly, we obtained the quality evaluation results, which are quite close to the original separate state-of-the-art DNNs and even outperforms them in some evaluation indices.

The Driving Assistance Systems (DAS) aim to help the vehicle drivers to proceed through different road situations. However, their main task is not only to safe one particular driver, but also to increase the safety for all traffic members. The problem domain here is huge and might be divided into the subtopics, like driverpsilas fatigue detection, pedestrian tracking, obstacle collision avoidance, lane departure warnings and traffic signs detection and recognition. Advanced computer vision techniques are widely used in order to develop sufficient and robust systems for driving assistance. In this paper we discuss the video-based Hough-Transform driven objects detection algorithms and their applications for lane departure warnings, as well as for traffic signs detection. Furthermore, a high-speed hardware implementation of these algorithms on the FPGA/ASIC is also presented.

The development of advanced driver assistance systems (ADAS) is of importance to prevent the potential driving risks on the road, which relies on the perceptive sensors to detect and identify the traffic conditions. This paper examines a method to track multiple objects and determine the high risk obstacles using radar. We experiment the scenarios as the vehicle and pedestrians in front, as well as the pedestrian crossing the street. The multiple tracks are achieved based on the Kalman filter. This work implements the state prediction model to compensate the missing targets, and the false positives are excluded through a threshold of bad tracks. To identify the potential obstacles, we impose two boundaries, i.e., lane and dynamic boundaries as the region of interest on the multiple tracks. It is demonstrated that the dynamic boundary benefits the finding of high risk obstacles and excludes most of the background noises. The lane boundary takes the advantage of tracking objects along the driving path: however, it could underestimate the risk of transverse moving objects in which the tracks are not continuous crossing the lanes.

This study examines the potential effectiveness of Lane Departure Warning (LDW) systems, an active safety system that aims to prevent or mitigate unintended departures from the lane while driving. Real-world departure crashes were extracted from a nationally representative sample of collisions in the U.S. from 2007 to 2011. The 2,848 collisions corresponded to 741,646 crashes. Because driver characteristics, e.g. steering input and reaction times, are unknown in retrospective crash databases, multiple simulations were performed for each crash, each with different driver characteristics. In addition, we developed a statistical framework for estimating initial vehicle conditions, i.e. departure speed, departure angle, road radius of curvature, and shoulder width. Trajectories were simulated to estimate the expected number of crashes and injuries that could be prevented by LDW. The expected reduction in road departure crashes due to LDW is estimated to be 30.3%. The expected reduction in injuries that are caused by LDW applicable crashes is estimated to be 25.8%.

Effects of an integrated collision warning system on teenage driver behavior

The detection of lane boundaries on suburban streets using images obtained from video constitutes a challenging task. This is mainly due to the difficulties associated with estimating the complex geometric structure of lane boundaries, the quality of lane markings as a result of wear, occlusions by traffic, and shadows caused by road-side trees and structures. Most of the existing techniques for lane boundary detection employ a single visual cue and will only work under certain conditions and where there are clear lane markings. Also, better results are achieved when there are no other on-road objects present. This paper extends our previous work and discusses a novel lane boundary detection algorithm specifically addressing the abovementioned issues through the integration of two visual cues. The first visual cue is based on stripe-like features found on lane lines extracted using a two-dimensional symmetric Gabor filter. The second visual cue is based on a texture characteristic determined using the entropy measure of the predefined neighbourhood around a lane boundary line. The visual cues are then integrated using a rule-based classifier which incorporates a modified sequential covering algorithm to improve robustness. To separate lane boundary lines from other similar features, a road mask is generated using road chromaticity values estimated from CIE L * a * b * colour transformation. Extraneous points around lane boundary lines are then removed by an outlier removal procedure based on studentized residuals. The lane boundary lines are then modelled with Bezier spline curves. To validate the algorithm, extensive experimental evaluation was carried out on suburban streets and the results are presented.

An important module in driver assistance systems is the lane departure warning which warns the driver when the vehicle begins to cross out of its lane inadvertently. A novel lane detection algorithm designed for lane departure warning system is presented in this paper. The proposed approach exploits the advantages offered by tree search to set up an intuitive and lightweight lane detection solution compared to existing ones. The algorithm is capable to work with poor light conditions, e.g., tunnel road, irrelevant markings on the road and the inherent long and thin property of lanes, scenarios with night conditions. Furthermore, the algorithm has a low time complexity which allows to be used on small and cheap hardware systems. An experimental setup containing a low cost front-facing camera sensor and an embedded computing board are used to implement the algorithm. A testing in a real highway scenario is carried out to validate the effectiveness of the proposed algorithm.