Explore the vast range of titles available.

Electric vehicle headlamp illuminance directly affects the driver’s visibility. Accurately predicting electric vehicle headlamp illuminance is crucial to enhancing driving safety. Existing deep learning models are trained using data collected from real-world road testing, yet external factors may compromise its reliability. Electric vehicle headlamp illuminance prediction primarily relies on data fitting, and such models are prone to overfitting when input data are affected by external disturbances. To solve the problem, we propose a luminancxel properties physical information neural network (LumiPINN) prediction model. Test conditions are designed in accordance with standard. The data was collected in an indoor laboratory to eliminate the influence of external factors, then underwent cleaning and pre-processing to ensure data quality. During the modelling process, the physical model is treated as a constraint, with the loss function to jointly optimise the prediction model. Compared with Deep Neural Network and Artificial Neural Network prediction models, the Mean Absolute Error, Mean Square Error, Root Mean Square Error, Mean Relative Error were reduced by 60.2%, 83.6%, 59.6%, 61.3%, and 71.7%, 90.7%, 69.5%, 71.4%. The Coefficient of Determination improved by 0.0015 and 0.0029. The results show that the LumiPINN prediction model demonstrates higher accuracy in prediction outcomes.

High-resolution light distributions are lately in demand for vehicle headlamp systems as an innovative lighting approach. This lighting approach can realize functionalities, such as precise glare avoidance and on-road projection, which are useful for improving traffic comfort and safety. For achieving the required high-resolution light distribution, area-based projection technologies, such as DMD, LCD, and LCoS, are considered to be integrated into such headlamps. These projection devices demand rectangular illumination areas with specific light distributions to fulfill the requirements for illumination efficiency and performance in headlamp systems. Lenslet arrays, based on the principle of Köhler illumination, can effectively homogenize the light and shape it into rectangular shapes simultaneously. Such components are widely used in projection applications. However, they also show functional potentialities to be applied in high-resolution headlamps. This paper explains the design principles and methods of lenslet arrays for beam pre-shaping in headlamp systems. It validates the homogenization using a self-designed and manufactured lenslet array in a demonstrator in the first place. Afterward, this paper introduces two new methods for the centralized beam shaping required by some headlamps. These methods are validated by optical simulations.

As one external lighting source on the road, headlamps from adjacent vehicles in the stream traffic should not be ignored. No comprehensive study has yet been developed for exploring the influence of sign luminance produced by other vehicle headlamps. In this paper, a luminance calculation model is developed to calculate sign luminance from all potential headlamps in the stream traffic. Using the model, four main scenarios have been simulated to analyze the effects of the positions of the target vehicle and other vehicles, vehicle type, sign type and sheeting material on the sign luminance. In addition, occlusion between vehicles is also addressed in the paper, by calculating the minimum distances between vehicles for the headlamps and for the driver's view of the following vehicle when vehicles and the sign are and are not in the same lane.

DEFT (Dynamic Environmental Filtering and Thresholding) algorithm is proposed to optimize vehicle height control and improve the performance of Adaptive Headlamp Systems. The DEFT algorithm is designed to enhance the reliability and stability of headlamp control systems by dynamically adapting to various driving conditions and environmental changes in real time. To detect and stabilize irregular fluctuations in vehicle height due to load variations, this study integratesreal-time data acquisition based on Hall sensors, dynamic boundary setting using linear interpolation, and signal stabilization by applying Kalman and Median filters with hysteresis. These components work together to suppress control signal instability caused by noise and abnormal signals, thereby reinforcing the reliability and consistency of vehicle height control. In particular, the hysteresis function reduces unnecessary signal fluctuations near threshold values, which not only extends the control system’s lifespan but also ensures stable operation. Experimental results demonstrate that the DEFT algorithm overcomes the limitations of conventional variable-resistance sensors and significantly enhances adaptive headlamp control performance under various driving conditions. This study presents a high-reliability solution for real-time vehicle height adjustment within ADASs (Advanced Driver Assistance Systems) and demonstrates the potential for application in diverse vehicle control systems.

With the emergence of LED lighting, IT convergence technology using the visible spectrum of LEDs, such as Visible Light Communication (VLC), has been highlighted. Among the many VLC applications, vehicular VLCs based on LED headlamps and transportation lighting infrastructure, such as street lamps, traffic lights, etc., are considered good alternatives for Intelligent Transportation Systems (ITS) or Active Safety applications. This paper introduces a demonstration system of vehicle-to-vehicle (V2V) VLC based on LED headlamps. By applying an inverse 4-PPM modulation scheme satisfying a 75 % dimming level under the light distribution regulation of LED headlamp, the proposed system showed its capability for V2V VLC with a 10 kbps data rate for more than 30 m under day time conditions. By measuring the BER performance according to distance, outdoor V2V VLC was possible for more than 30 m even in the day time.

In recent years, computer vision algorithms have become more powerful, which enabled technologies such as autonomous driving to evolve rapidly. However, current algorithms mainly share one limitation: They rely on directly visible objects. This is a significant drawback compared to human behavior, where visual cues caused by objects (e. g., shadows) are already used intuitively to retrieve information or anticipate occurring objects. While driving at night, this performance deficit becomes even more obvious: Humans already process the light artifacts caused by the headlamps of oncoming vehicles to estimate where they appear, whereas current object detection systems require that the oncoming vehicle is directly visible before it can be detected. Based on previous work on this subject, in this paper, we present a complete system that can detect light artifacts caused by the headlights of oncoming vehicles so that it detects that a vehicle is approaching providently (denoted as provident vehicle detection). For that, an entire algorithm architecture is investigated, including the detection in the image space, the three-dimensional localization, and the tracking of light artifacts. To demonstrate the usefulness of such an algorithm, the proposed algorithm is deployed in a test vehicle to use the detected light artifacts to control the glare-free high beam system proactively (react before the oncoming vehicle is directly visible). Using this experimental setting, the provident vehicle detection system’s time benefit compared to an in-production computer vision system is quantified. Additionally, the glare-free high beam use case provides a real-time and real-world visualization interface of the detection results by considering the adaptive headlamps as projectors. With this investigation of provident vehicle detection, we want to put awareness on the unconventional sensing task of detecting objects providently (detection based on observable visual cues the objects cause before they are visible) and further close the performance gap between human behavior and computer vision algorithms to bring autonomous and automated driving a step forward.

In this study, we present an algorithm to estimate the distance between a vehicle and a target object using light from headlights captured by a camera. In situations with limited distance data, we also design a fuzzy controller using the adaptive neuro–fuzzy inference system (ANFIS). To enhance robustness against disturbances, the interval type-2 approach is used. For the distance estimation algorithm, the vehicle is positioned at predefined intervals from the target object, capturing images of the headlights at each point. The region of interest containing the light is extracted from each image and segmented by light intensity. Weighted values are then assigned to each segment based on intensity, producing an image value that correlates with the distance. This image-derived value is then used as distance data for the design of the fuzzy controller. The controller is implemented using the interval type-2 fuzzy logic toolbox in MATLAB/SIMULINK, with vehicle speed and image intensity values as inputs and control torque as the output to adjust vehicle speed. The noise from the vehicle speed sensor is treated as a disturbance, and the performance of the interval type-2 fuzzy controller is evaluated under these disturbance conditions. Additionally, fuzzy controllers are designed for vehicle positions between 41–43 m and 47–49 m, and these controllers are trained using ANFIS to function effectively across the entire 41–49 m range. Simulation results demonstrate that, with the controller integrated into the vehicle system, the vehicle is successfully controlled to reach the target position.

Driver assistance systems or smart headlamp technology require accurate vehicle detection in nighttime traffic environments. This study proposes a novel nighttime vehicle detection (NVD) algorithm that can be applied to these technologies, and the algorithm was implemented using image processing. Images taken on roads at night are basically dark and lack information regarding vehicle appearance. Additionally, they contain significant noise caused by various lights and the scattering or reflection of these lights. Therefore, it is difficult to increase the performance of existing NVD methods. In addition, recent end-to-end convolutional neural network (CNN)-based object detection (OD) methods exhibit low NVD performance owing to their poor learning capability caused by lack of information and noise in the images. This study presents new methods to overcome the limitations of NVD implementation: 1) We propose a candidate search and diffusion method based on the use of experimental heuristics and hand-crafted features to utilize the characteristics related to light emitted from vehicle headlamps or taillamps. 2) We propose a CNN method that combines the approaches applied in latest CNN-based OD method with the proposed candidate search and diffusion method. To demonstrate the superiority of the proposed method, we conducted experiments and compared the proposed method to recent CNN-based OD methods. The experimental results demonstrated the higher detection performance of the proposed method compared to other methods. The code is available at https://github.com/sjg918/gj-nvd-diffusion/

In the visible light communication (VLC) for parallel vehicles, the light emitted by the headlamps cannot reach the photodiode (PD), which is the receiver installed on another vehicle. To solve this problem, the mirror array-based intelligent reflecting surface- (IRS-) aided VLC system is designed for parallel vehicles, and the system performance with different number of mirrors and transmission distance is analyzed. The maximum distance between adjacent IRSs is calculated via the exhaustive method, and the nearest neighbor iterative search (NNIS) algorithm is proposed for IRS node selection. Numerical results show that the SNR increases with the number of mirrors increasing. If the number of mirrors is 6×6 in the IRS, the maximum distance between adjacent IRSs is 37 meters. When the interval between adjacent IRSs is 32 meters, the required BER can be satisfied with merely three IRSs working together according to the NNIS algorithm.

The use of mobile laser scanning for mapping forests has scarcely been studied in difficult forest conditions. In this paper, we compare the accuracy of retrieving tree attributes, particularly diameter at breast height (DBH), stem curve, stem volume, and tree height, using six different laser scanning systems in a managed natural boreal forest. These compared systems operated both under the forest canopy on handheld and unmanned aerial vehicle (UAV) platforms and above the canopy from a helicopter. The complexity of the studied forest sites ranged from easy to difficult, and thus, this is the first study to compare the performance of several laser scanning systems for the direct measurement of stem curve in difficult forest conditions. To automatically detect tree stems and to calculate their attributes, we utilized our previously developed algorithm integrated with a novel bias compensation method to reduce the overestimation of stem diameter arising from finite laser beam divergence. The bias compensation method reduced the absolute value of the diameter bias by 55–99%. The most accurate laser scanning systems were equipped with a Velodyne VLP-16 sensor, which has a relatively low beam divergence, on a handheld or UAV platform. In easy plots, these systems found a root-mean-square error (RMSE) of below 10% for DBH and stem curve estimates and approximately 10% for stem volume. With the handheld system in difficult plots, the DBH and stem curve estimates had an RMSE under 10%, and the stem volume RMSE was below 20%. Even though bias compensation reduced the difference in bias and RMSE between laser scanners with high and low beam divergence, the RMSE remained higher for systems with a high beam divergence. The airborne laser scanner operating above the forest canopy provided tree attribute estimates close to the accuracy of the under-canopy laser scanners, but with a significantly lower completeness rate for stem detection, especially in difficult forest conditions.