Driving Assistance Systems such as lane departure and front collision warning has caught great attention for its promising usage on road driving. This, this research focus on implementing lane departure and front collision warning at same time. In order to make the system really useful for real situation, it is critical that the whole process could be near real-time. Thus we chose Hough Transform as the main algorithm for detecting lane on the road. Hough Transform is used for that it is a very fast and robust algorithm, which makes it possible to execute as many frames as possible per frames. Hough Transform is used to get boundary information, so that we could decide if the car is doing lane departure based on the car's position in lane. Later, we move on to use front car's symmetry character to do front car detection, and combine it with Camshift tracking algorithm to fill the gap for failure of detection. Later we introduce camera calibration, stereo calibration, and how to calculate real distance from depth map.

Cross-centerline crashes occur rarely in the United States but are especially severe. This type of crash is characterized by one vehicle departing over a centerline and encountering a vehicle traveling in the opposite direction. In recent years, automakers have started developing and implementing lane departure warning (LDW) on newer vehicles. This system provides the potential to reduce or significantly impact the frequency of cross-centerline crashes. The objective of this thesis was to estimate the potential crash and injury benefits of a LDW system if installed on every vehicle in the US fleet. This research includes the following 1) a characterization of cross-centerline crashes in the United States today with current and future prevention methods, 2) a reconstruction methodology used for all crashes including rollovers and heavy vehicles, and 3) a simulation model and approach method used to estimate potential benefits of LDW systems on cross-centerline crashes. Cross over to left crashes account for only 4% of non-junction non-interchange crashes but account for 44% of serious injury crashes of the same type. As part of this research, 42 cross-centerline crashes were reconstructed and simulated as if they had a LDW system installed. Accounting for driver capability to react to a LDW alert, crash reduction benefits ranged from 22 – 30%.Using injury risk curves, the probability of experiencing a MAIS2+ injury in a cross-centerline crash was reduced by 29% when using a LDW system. / Master of Science

<p>A lane departure warning system relying exclusively on a camera has several shortcomings and tends to be sensitive to, e.g., bad weather and abrupt manoeuvres. To handle these situations, the system proposed in this thesis uses a dynamic model of the vehicle and integration of relative motion sensors to estimate the vehicle’s position on the road. The relative motion is measured using vision, inertial, and vehicle sensors. All these sensors types are affected by errors such as offset, drift and quantization. However the different sensors are sensitive to different types of errors, e.g., the camera system is rather poor at detecting rapid lateral movements, a type of situation which an inertial sensor practically never fails to detect. These kinds of complementary properties make sensor fusion interesting. The approach of this Master’s thesis is to use an already existing lane departure warning system as vision sensor in combination with an inertial measurement unit to produce a system that is robust and can achieve good warnings if an unintentional lane departure is about to occur. For the combination of sensor data, different sensor fusion models have been proposed and evaluated on experimental data. The models are based on a nonlinear model that is linearized so that a Kalman filter can be applied. Experiments show that the proposed solutions succeed at handling situations where a system relying solely on a camera would have problems. The results from the testing show that the original lane departure warning system, which is a single camera system, is outperformed by the suggested system.</p>

A lane departure warning system relying exclusively on a camera has several shortcomings and tends to be sensitive to, e.g., bad weather and abrupt manoeuvres. To handle these situations, the system proposed in this thesis uses a dynamic model of the vehicle and integration of relative motion sensors to estimate the vehicle’s position on the road. The relative motion is measured using vision, inertial, and vehicle sensors. All these sensors types are affected by errors such as offset, drift and quantization. However the different sensors are sensitive to different types of errors, e.g., the camera system is rather poor at detecting rapid lateral movements, a type of situation which an inertial sensor practically never fails to detect. These kinds of complementary properties make sensor fusion interesting. The approach of this Master’s thesis is to use an already existing lane departure warning system as vision sensor in combination with an inertial measurement unit to produce a system that is robust and can achieve good warnings if an unintentional lane departure is about to occur. For the combination of sensor data, different sensor fusion models have been proposed and evaluated on experimental data. The models are based on a nonlinear model that is linearized so that a Kalman filter can be applied. Experiments show that the proposed solutions succeed at handling situations where a system relying solely on a camera would have problems. The results from the testing show that the original lane departure warning system, which is a single camera system, is outperformed by the suggested system.

In lane departure warning systems (LDWS) it is important that the auditory warning triggers a fast and appropriate reaction from the driver. The rumble strip noise is a suitable warning to alert the driver of an imminent lane departure. A short reaction time is important in lane departure scenarios, where a late response may have fatal consequences. For abstract sounds an increase in intensity can influence the perceived urgency level of the warning, which may also trigger a faster reaction from the listener. In this thesis, the effect of a rumble strip warning with looming (increasing) intensity was analyzed by letting test persons drive a driving simulator and measuring how quickly they reacted to the auditory warning. These results were compared with those for a rumble strip warning with a constant intensity, and two versions of an abstract warning; constant intensity and looming intensity. A survey regarding the perceived urgency, annoyance and acceptance of the warnings was also carried out. The results show no differences in reaction time between the four warning signals. This may be because the test persons expected the warnings, or because of their limited experience. The survey suggests that adding a looming intensity to the rumble strip warning results in a higher urgency, while keeping the annoyance low, which could be of importance to avoid unwanted reactions from the driver. / I varningssystem för personbilar används ofta ett system som signalerar ett stundande ofrivilligt lämnande av körfältet, s.k. lane departure warning systems (LDWS), genom att en varningssignal ljuder. Det är viktigt att en sådan akustisk varningssignal frammanar en snabb och lämplig reaktion från föraren. Ljudet av en bullerräffla är en lämplig varningssignal för detta ändamål. En kort reaktionstid är viktig när fordon är på väg att ofrivilligt lämna körfältet, då en långsam reaktion kan ha förödande konsekvenser. Studier på abstrakta akustiska varningssignaler har visat att en ökande intensitet kan få en varning att verka mer brådskande, vilket i sin tur kan leda till att lyssnaren reagerar snabbare. I denna rapport analyseras hur ett bullerräffleljuds ökande intensitet påverkar förarens reaktionstid. Analysen gjordes genom att mäta reaktionstiden hos testpersoner som körde en bilsimulator med fyra olika varningssignaler; en bullerräffleljudsvarning och en abstrakt varning, båda med konstant intensitet och ökande intensitet. Reaktionstiderna för de olika signalerna jämfördes, varpå en enkät utfärdades där testpersonerna uppgav hur brådskande och irriterande de uppfattade varningarna, samt till vilken grad de skulle acceptera varningarna i ett verkligt körscenario. Resultaten visar inga skillnader i reaktionstid mellan varningarna, vilket kan bero på att testpersonerna förutsåg när varningarna skulle komma, eller på grund av deras begränsade erfarenhet av bullerräffleljud. Enkätens utfall antyder att bullerräffleljudsvarningen med ökande intensitet är mer brådskande än versionen med konstant intensitet, men att irritationsnivån inte påverkas när intensiteten ökar, vilket kan vara viktigt för att inte framkalla oönskade reaktioner hos föraren.

Indiana University-Purdue University Indianapolis (IUPUI) / In the past decades, the rapid innovation of technology has greatly affected the automotive industry. However, every innovation has always been paired with safety risks that needs to be quickly addressed. This is where Petri nets (PNs) have come into the picture and have been used to model complex systems for different purposes, such as production management, traffic flow estimation and the introduction of new car features collectively known as, Adaptive Driver Assistance Systems (ADAS). Since most of these systems include both discrete and continuous dynamics, the Hybrid Petri net (HPN) model is an essential tool to model these. The objective of this thesis is to develop, analyze and simulate a lane keeping support system using an HPN model. Chapter 1 includes a brief summary of the specific ADAS used, lane departure warning and lane keeping assist systems and then related work on PNs is mentioned. Chapter 2 provides a background on Petri nets. In chapter 3, we develop a discrete PN model first, then we integrate continuous dynamics to extend it to a HPN model that combines the functionalities of the two independent ADAS systems. Several scenarios are introduced to explain the expected model behavior. Chapter 4 presents the analysis and simulation results obtained on the final model. Chapter 5 provides a summary for the work done and discusses future work.

Doctor of Philosophy / Department of Civil Engineering / Sunanda Dissanayake / Lane-departure crashes are the most predominant crash type in Kansas which causes very high number of motor vehicle fatalities. Therefore, the Kansas Department of Transportation (KDOT) has implemented several different types of countermeasures to reduce the number of motor vehicle fatalities associated with lane-departure crashes. This research was conducted to estimate the safety effectiveness of commonly used lane-departure countermeasures in Kansas on all crashes and lane-departure crashes using Crash Modification Factors (CMFs). Paved shoulders, rumble strips, safety edge treatments and median cable barriers were identified as the commonly used lane-departure countermeasures on both tangent and curved road segments while chevrons and post-mounted delineators were identified as the most commonly used lane-departure countermeasures on curved road segments. This research proposes a state-of-art method of estimating CMFs using cross-sectional data for chevrons and post-mounted delineators. Furthermore, another state-of-art method is proposed in this research to estimate CMFs for safety edge treatments using before-and-after data. Considering the difficulties of finding the exact date of implementation of each countermeasure, both cross-sectional and before-and-after studies were employed to estimate the CMFs. Cross-sectional and case-control methods, which are the two major methods in cross-sectional studies were employed to estimate CMFs for paved shoulders, rumble strips, and median cable barriers. The conventional cross-sectional and case-control methods were modified when estimating CMFs for chevrons and post-mounted delineators by incorporating environmental and human behaviors in addition to geometric and traffic-related explanatory variables. The proposed method is novel and has not been used in the previous cross-sectional models available in the literature. Generalized linear regression models assuming negative binomial error structure were used to develop models for cross-sectional method to estimate CMFs while logistic regression models were used to estimate CMFs using case-control method. Results showed that incorporating environmental and human-related variables into cross-sectional models provide better model fit than in conventional cross-sectional models. To validate the developed models for cross-sectional method, mean of the residuals and the Root Mean Square Error (RMSE) were used. For the case-control method, Receiver Operational Characteristic (ROC) was used to evaluate the predictive power of models for a binary outcome using classification tables. However, it was seen that the case-control method is not suitable for estimating CMFs for all crashes since the range of the crash frequency is wide in each road segment. A regression-based method of estimating CMFs using before-and-after data was proposed to estimate CMFs for safety edge treatments. This method allows researchers to identify the safety effectiveness of an individual CMFs on road segments where multiple treatments have been applied at the same time. Since this method uses road geometric and traffic-related characteristics in addition to countermeasure information as the explanatory variables, the model itself would be the Safety Performance Function (SPF). Therefore, developing new SPF is not necessary. Finally, the CMFs were estimated using before-and-after Empirical Bayes method to validate the results from the regression-based method. The results of this study can be used as a decision-making tool when implementing lane-departure countermeasures on similar roadways in Kansas. Even though there are readily available CMFs from the national level studies, having more localized CMFs will be beneficial due to differences in traffic-related and geometric characteristics on different roadways.

With the rapid improvement of the Advanced Driver Assistant System (ADAS), autonomous driving has become one of the most common hot topics in recent years. While driving, many technologies related to autonomous driving choose to use the sensors installed on the vehicle to collect the information of road status and the environment outside. This aims to warn the driver to perceive the potential danger in the fastest time, which has become the focus of autonomous driving in recent years. Although autonomous driving brings plenty of conveniences to people, the safety of it is still facing difficulties. During driving, even the experienced driver can not guarantee focus on the status of the road all the time. Thus, lane departure warning system (LDWS) becomes developed. The purpose of LDWS is to determine whether the vehicle is in the safe driving area. If the vehicle is out of this area, LDWS will detect it and alert the driver by the sensors, such as sound and vibration, in order to make the driver back to the safe driving area. This thesis proposes a novel lightweight LDWS model LEHA, which divides the entire LDWS into three stages: image preprocessing, lane detection, and lane departure recognition. Different from the deep learning methods of LDWS, our LDWS model LEHA can achieve high accuracy and efficiency by relying only on simple hardware. The image preprocessing stage aims to process the original road image to remove the noise which is irrelevant to the detection result. In this stage, we apply a novel algorithm of grayscale preprocessing to convert the road image to a grayscale image, which removes the color of it. Then, we design a binarization method to greatly extract the lane lines from the background. A newly-designed image smoothing is added to this stage to reduce most of the noise, which improves the accuracy of the following lane detection stage. After obtaining the processed image, the lane detection stage is applied to detect and mark the lane lines. We use region of interest (ROI) to remove the irrelevant parts of the road image to reduce the detection time. After that, we introduce the Canny edge detection method, which aims to extract the edges of the lane lines. The last step of LDWS in the lane detection stage is a novel Hough transform method, the purpose of it is to detect the position of the lane and mark it. Finally, the lane departure recognition stage is used to calculate the deviation distance between the vehicle and the centerline of the lane to determine whether the warning needs to turn on. In the last part of this paper, we present the experiment results which show the comparison results of different lane conditions. We do the statistic of the proposed LDWS accuracy in terms of detection and departure. The detection rate of our proposed LDWS is 98.2% and the departure rate of it is 99.1%. The average processing time of our proposed LDWS is 20.01 x 10⁻³s per image.

Every year, approximately 34,000 individuals are fatally injured in crashes on US roads [1]. These fatalities occur across many types of crash scenarios, each with its own causation factors. One way to prioritize research on a preventive technology is to compare the number of occupant fatalities relative to the total number of occupants involved in a crash scenario. Four crash modes are overrepresented among fatalities: single vehicle road departure crashes, control loss crashes, cross-centerline head-on crashes, and pedestrian/cyclist crashes [2]. Interestingly, three of these crash scenarios require the subject vehicle to depart from the initial lane of travel. Lane departure warning (LDW) systems track the vehicle lane position and can alert the driver through audible and haptic feedback before the vehicle crosses the lane line. Lane departure prevention (LDP) systems can perform an automatic steering maneuver to prevent the departure. Another method of prioritizing research is to determine factors common among the fatal crashes. In 2017, 30.4% of passenger vehicle crash fatalities involved a vehicle rollover [1]. Half of all fatal single vehicle road departure crashes resulted in a rollover yet only 12% of fatal multi-vehicle crashes involved a rollover [1]. These often occur after the driver has lost control of the vehicle and departed the road. Electronic stability control (ESC) can provide different braking to each wheel and allow the vehicle to maintain heading. While ESC is a promising technology, some rollover crashes still occur. Passive safety systems such as seat belts, side curtain airbags, and stronger roofs work to protect occupants during rollover crashes. Seat belts prevent occupants from moving inside the occupant compartment during the rollover and both seat belts and side curtain airbags can prevent occupants from being ejected from the vehicle. Stronger roofs ensure that the roof is not displaced during the rollover and the integrity of the occupant compartment is maintained to prevent occupant ejection. The focus of this dissertation is to evaluate the effectiveness of vehicle-based countermeasures, such as lane departure warning and electronic stability control, for preventing or mitigating single vehicle road departure crashes, cross-centerline head-on crashes, and single vehicle rollover crashes. This was accomplished by understanding how drivers respond to both road departure and cross-centerline events in real-world crashes. These driver models were used to simulate real crash scenarios with LDW/LDP systems to quantify their potential crash reduction. The residual crashes, which are not avoided with LDW/LDP systems or ESC, were analyzed to estimate the occupant injury outcome. For rollover crashes, a novel injury model was constructed that includes modern passive safety countermeasures such as seat belts, side curtain airbags, and stronger roofs. The results for road departure, head-on, and control loss rollover crashes were used to predict the number of crashes and injured occupants in the future. This work is important for identifying the residual crashes that require further research to reduce the number of injured crash occupants. / Doctor of Philosophy / Every year in the US, approximately 34,000 individuals are fatally injured in many different types of crashes. However, some crash types are more dangerous than other crash types. Drift-out-of-lane (DrOOL) road departure crashes, control loss road departure crashes, head-on crashes, and pedestrian crashes are more likely to result in an occupant fatality than other crash modes. In three of these more dangerous crash types, the vehicle departs from the lane before the crash occurs. Lane departure warning (LDW) systems can detect when the vehicle is about to cross the lane line and notify the driver with beeping or vibrating the steering wheel. A different system, called lane departure prevention (LDP), can provide automatic steering to prevent the vehicle from leaving the lane or return lane. In control loss crashes, the vehicle's motion is in a different direction than the vehicle's heading. During control loss, it is easier for the vehicle to roll over which is very dangerous. Electronic stability control (ESC) can prevent control loss by applying selective braking to each tire to keep the vehicle's motion in the same direction as the vehicle's heading. If a rollover still occurs, vehicles are equipped with passive safety systems and designs such as seat belts, side curtain airbags, and stronger roofs to protect the people inside. Seat belts can prevent occupants from striking the vehicle interior during the rollover and both seat belts and side curtain airbags can prevent occupants from being ejected from the vehicle. Stronger roofs ensure that the roof is not displaced during the rollover to prevent occupants from being ejected from the vehicle. The focus of this dissertation is to estimate how many crashes LDW, LDP, and ESC systems could prevent. This was accomplished by understanding how drivers respond after leaving their lane in real crashes. Then, these real crash scenarios were simulated with an LDW or LDP system to estimate how many crashes were prevented. The occupants of residual crashes, which were not prevented by the simulated systems, were analyzed to estimate the number of occupants with at least one moderate injury. Understanding which crashes and injuries that were not prevented with this technology can be used to decide where future research should occur to prevent more fatalities in road departure, head-on and control loss crashes.

The objective of this dissertation is to develop a Multi-Camera Lane Departure Warning (MCLDW) system and a framework to evaluate it. A Lane Departure Warning (LDW) system is a safety feature that is included in a few luxury automobiles. Using a single camera, it performs the task of informing the driver if a lane change is imminent. The core component of an LDW system is a lane detector, whose objective is to find lane markers on the road. Therefore, we start this dissertation by explaining the requirements of an ideal lane detector, and then present several algorithmic implementations that meet these requirements. After selecting the best implementation, we present the MCLDW methodology. Using a multi-camera setup, MCLDW system combines the detected lane marker information from each camera's view to estimate the immediate distance between the vehicle and the lane marker, and signals a warning if this distance is under a certain threshold. Next, we introduce a procedure to create ground truth and a database of videos which serve as the framework for evaluation. Ground truth is created using an efficient procedure called Time-Slicing that allows the user to quickly annotate the true locations of the lane markers in each frame of the videos. Subsequently, we describe the details of a database of driving videos that has been put together to help establish a benchmark for evaluating existing lane detectors and LDW systems. Finally, we conclude the dissertation by citing the contributions of the research and discussing the avenues for future work.