Refine Your Search

Search Results

Technical Paper

Road Sign Recognition System Based on Wavelet Transform and OPSA point Set Distance

2018-08-07
2018-01-1609
Abstract: Signage recognition is one of the hot topics in recent years. It has important applications in intelligent traffic and autonomous driving of smart cars. This paper designs a road marking recognition method combining OPSA point set distance and wavelet transform. The method consists of three main phases: 1) image denoising, restora-tion, 2) feature extraction, and 3) image recognition. First, a Gauss-ian-smoothing filter used to attenuate or remove irrelevant information in the image, enhance related information in the image, and achieve image denoising. In the feature extraction stage, the feature extraction and recognition method based on wavelet transform adopted to overcome the deficiency of the traditional Fourier feature extraction method, ensure that high frequency information is not lost, and low frequency information is not lost. Finally, the OSPA point set used to identify distance markers.
Technical Paper

A Real-Time Traffic Light Detection Algorithm Based on Adaptive Edge Information

2018-08-07
2018-01-1620
Traffic light detection has great significant for unmanned vehicle and driver assistance system. Meanwhile many detection algorithms have been proposed in recent years. However, traffic light detection still cannot achieve a desirable result under complicated illumination, bad weather condition and complex road environment. Besides, it is difficult to detect multi-scales traffic lights by embedded devices simultaneously, especially the tiny ones. To solve these problems, this paper presents a robust vision-based method to detect traffic light, the method contains main two stages: the region proposal stage and the traffic light recognition stage. On region proposal stage, we utilize lane detection to remove partial background from the original images. Then, we apply adaptive canny edge detection to highlight region proposal in Cr color channel, where red or green color proposals can be separated easily. Finally, extract the enlarged traffic light RoI (Region of Interest) to classify.
Technical Paper

Automatic Azimuth Alignment for Automotive Radar

2018-08-07
2018-01-1606
The world has witnessed the rapid development of the Advanced Driver Assist System (ADAS) industry over the past few years. Radar, as one of the most important sensors in ADAS due to the high penetration, all-weather characteristic and low cost, is studied intensively and will be applied on a large scale. Automobile radar has many applications like ACC (Advanced Cruise Control) BSD (Blind Spot Detection), LCA (Lane Change Assistant), etc., and the accuracy of the radar target detection influences the performance of ADAS. In general, range, velocity, azimuth angle and other target attributions are measured by the automotive radar, and the accuracy of the azimuth angle is more easily affected by the environment than other attributions. For the automotive radar, it is usually equipped either near a front bumper, or near a left rear and right rear bumper.
Technical Paper

System design and Model of a 3D 79GHz High Resolution Ultra-wide band millimeter-wave imaging automotive radar

2018-08-07
2018-01-1615
Automotive radar is an important environment perception sensor for advance driving assistance system. It can detect objects around the vehicle with high accuracy and it works in all bad weathers. For traditional automotive radar, it cannot measure the objects’ height. Thus, a manhole cover on the road surface or a guideboard high above the road would be taken erroneously as a non-moving car. In such cases, the adaptive cruise system would decelerate or stop the vehicle erroneously and make the driver uncomfortable. A 3D automotive radar with two-dimensional electronic scanning can measure the targets’ height as well as the targets’ azimuth angle. This paper presents a 79GHz ultra-wide band automotive 3D imaging radar. Due to the 4 GHz wide bandwidth, the range resolution of this radar can be as small as 3.75 cm.
Technical Paper

Evaluation and Optimization of Driver Steering Override Strategy for LKAS Based on Driver’s Acceptability

2018-08-07
2018-01-1631
In order to satisfy the design requirements of Lane Keeping Assistance System (LKAS), a driver steering override (DSO) strategy is necessary to deal with the driver’s interaction with the system. And the system can be overridden by the strategy in case of lane change, obstacle avoidance and some other emergencies. However, evaluation and optimization of the DSO strategy for LKAS cannot easily be completed quantitatively in consideration of driver’s acceptability due to the lack of a reliable process. In this research, firstly, subjective and objective evaluation experiment is designed. To be specific, for the objective evaluation experiment, several objective characteristic indices (CI) are extracted to describe the outward performance of DSO strategy such as the steering hand torque instead of using the parameters from a certain type of controller.
Technical Paper

Modified Car Following and Lane Changing Simulations Model for Autonomous Vehicle on Highway

2018-08-07
2018-01-1647
Being one of the most simple and basic driving scenarios, highway scenario can be one of the first scenarios to achieve autonomous driving. Both car following (CA) and lane changing (LC) are the most basic and frequent manoeuvre during highway driving task, and therefore become two key issues to focus on in recent researches about autonomous vehicle (AV). Different from conventional CA and LC researches that attach much importance to the character, psychology, perception ability, and driving experience of human drivers, more timely and accurate reactions based on fast perception and communication technology as well as the automatic actuator are hypotheses for this research. And based on these hypotheses, a modified intelligent driver model (MIDM) is proposed for AVs’ following behavior to alleviate the fluctuations caused by lane changing behaviors.
Technical Paper

Study on lane change trajectory planning considering of driver characteristics

2018-08-07
2018-01-1627
Automatic lane change of intelligent vehicles is a complex process. Besides of safety, feelings of the driver and passengers during the lane change are also very important. In this paper, a lane change trajectory planner is designed to generate an ideal collision-free trajectory to satisfy the driver's preference. Various lane changing modes, gentle lane change, general lane change, radical lane change and personalized lane change, are designed to meet the needs of different passengers on vehicles simultaneously. In this paper, the condition of the two-lane change is studied. One vehicle is in front of the ego vehicle at the same lane and one is at the rear of the ego vehicle at the target lane. A trajectory planning method is then established based on constant speed offset and sine curve, vehicle distances and speed difference, etc. The key factors which can reflect drivers’ lane change characteristics are then acquired.
Technical Paper

Mechanism analysis and simulation study of Automobile Millimeter Wave Radar Noise

2018-08-07
2018-01-1641
As we all know, millimeter wave radar has become one of the most important sensors of advanced driving assistance system (ADAS). However, though millimeter wave radar has strong environmental adaptability because of its unique frequency band, but its application performance is still largely affected by environmental noise, the main problem is that the environmental noise may cause the common problems such as false alarm, missing detection, inaccurate distance measurement, inaccurate velocity measurement, and inaccurate angle measurement, and so on. Therefore, R & D engineers must carry out a large number of repeated tests, it may develop a higher reliability, higher environmental adaptability of millimeter wave radar applications, extract the precise target data, and develop ADAS system more perfect. There are two ways to repeated tests, that is, the actual test and virtual test.
Technical Paper

Study on the Controlled Filed Test Scenarios of Automated Vehicles

2018-08-07
2018-01-1633
The controlled filed test is an essential part of evaluation methods to verify the safety, intelligence and comfort of automated vehicles. Due to the complex, dynamic and large masses of real-world traffic situations, it is very important to generate critical test scenarios with a limited number and a wide coverage, which could evaluate the function features and performance of automated vehicles at different intelligent levels. In this paper, we propose a method to form the basic driving condition group based on function features of level 2 and level 3 automated vehicles, through the permutation and combination of all possible relative position and movement relationships between ego-vehicle and surrounding vehicles in accordance with different traffic conditions and weather types.
Technical Paper

A Robust Path Tracking Control Method for Intelligent Vehicle

2018-08-07
2018-01-1582
Due to the nonlinearity and strong coupling between states of the intelligent vehicle, the influence of external disturbance and the complexity of driving conditions, the traditional path tracking control methods which are based on precise mathematical model are difficult to get a better control effect and cannot meet intelligent driving high-performance control requirements. This paper presents a strong robust path tracking control method which is based on sliding mode control and active disturbance rejection control and does not depend on accurate mathematical models. Firstly, by constructing a desired yaw angle function, which can guarantee that the deviations of the vehicle actual lateral displacement from the desired path can converges to zero when the yaw angle of the vehicle approaches the desired yaw angle, so that the complex path tracking control problem can be transformed into easy to implement yaw angle tracking control problem.
Technical Paper

A Path Planning Method Based on Large Amount r of Artificial Driving Trajectories

2018-08-07
2018-01-1588
Path planning plays an important role in autopilot technology, and its algorithm will have a key impact on the performance of autopilot system. The most commonly used path planning algorithms, such as A^* algorithm and Dijkstra algorithm, are based on the shortest distance principle. However, in real cases, when vehicle in obstacle avoidance or over bending scenes, ride comfort needs to be considered. To solve this problem, a path planning method based on large amount human driving trajectories is proposed in this paper. It enables autopilot system to learn human driving behavior and independently plan a more suitable method according to human driving habits. First step, we use map data and the artificial driving trajectories as the system inputs, and that provides an artificial driving trajectory database for path planning. Large amount of Artificial driving trajectories and high precision map insure that one or more paths can be plan out based on the human driving habits.
Technical Paper

Research on the Development Trend of Brain Controlled Cars

2018-08-07
2018-01-1587
A brain controlled car is an application of brain controlled technology in on-road motor vehicle. Driver can take control of a car by his electroencephalogram (EEG) to move forward, accelerate, as well as turn the steering wheel, and so on. The purposes of the studies on brain controlled car are in one hand to offer a completely new driving style and make people with physical disability to drive a car possible, and in the other hand to use the brain controlled technology as one of the advanced driver assistant methods, to make the driving experience more safe, more comfortable, more intelligent and more compliant to the driver’s intention. This paper summarizes the research status of brain controlled car technology based on both publicly released demo cars and academic articles. The car’s actions that brain controlled, the technology implementation roadmaps, and the major research directions are mainly researched.
Technical Paper

The Effect of Mounting Orientation of Resistive Particulate Matter Sensor on Signal Behavior

2018-08-07
2018-01-1598
Particulate matter (PM) sensors are required to monitor the diesel particulate filters (DPF) malfunction. The resistive PM sensor concept is widely chosen for this purpose due to functionality, costs and durability. The output signal of resistive PM sensor for interpreting and processing for diagnosing DPF status is significantly affected by the exhaust velocity and soot concentration in the vicinity of sensor sensing element. Theoretically, During the regeneration of PM sensor, no new accumulation of particles is possible. Even after the regeneration, PM cannot immediately be accumulated again, because of thermal inertia, the PM sensor requires a certain time for the thermalization of the sensor element by the exhaust gas. During the regeneration phase and the subsequent cooling phase,PM sensor is insensitive with respect to a possibly present soot concentration , but the choking phenomena of PM sensor can be observed, when soot depositing rate reaches a threshold value.
Technical Paper

Lateral control method of intelligent vehicles based on image segmentation

2018-08-07
2018-01-1596
With the rapid development of automobile industry, the intelligent vehicles that can be viewed as the integrated carrier of advanced technology of automobile are paid much attention to by society. It is imperative to study the motion control of the intelligent vehicles due to the nature of their nonholonomic operation constraint system whose dynamic characteristics are highly nonlinear with the uncertainty of parameters. In this paper, utilizing the vision system of intelligent vehicles, a vehicle lateral control strategy based on image segmentation is established to enhance the vehicle's capability to predict future behavior and deal with unexpected situations. Applying the image recognition and tracking results of the visual system, the breadth and depth of the vision are divided into three-dimensional segmentation where each block gives different weights.
Technical Paper

Semantic Segmentation for Traffic Scene Understanding based on Mobile Networks

2018-08-07
2018-01-1600
Real-time and reliable perception of the surrounding environment is an important prerequisite for advanced driving assistance system (ADAS) and automatic driving. And vision-based detection plays a significant role in environment perception for automatic vehicles. Although deep convolutional neural networks enable efficient recognition of various objects, it has difficulty in accurately detecting special vehicles, rocks, road pile, construction site, fence and so on. In this work, we address the task of traffic scene understanding with semantic image segmentation. Both driveable area and the classification of object can be attained from the segmentation result. First, we define 29 classes of objects in traffic scenarios with different labels and modify the Deeplab V2 network. Then in order to reduce the running time, MobileNet architecture is applied to generate the feature map instead of the original models.
Technical Paper

Decade of Vision-based Pedestrian Detection for Self-driving: An Experimental Survey and Evaluation

2018-08-07
2018-01-1603
With the steady progress in autonomous driving technology and the tremendous potential prospects for development, the topics about self-driving car have begun return to the center stage of AI applications. Hence, lots of efforts are made on algorithms relevant to self-driving car itself. However, few shed the light on the problem of how to testify them thoroughly, therefore unified standards for autonomous driving and testing is urgently needed. To study this problem, we begin by pedestrian detection, for that the ability of locating humans is one of the most critical problems that should be concerned about for self-driving cars. In this paper, we investigate to perform a standard evaluation to qualify different methods and detectors under a more practical manner. Specifically, we investigate several commonly used evaluation methodologies for pedestrian detection, and find out that the Caltech pedestrian detection benchmark is the most popular.
Technical Paper

Camera-Radar Data Fusion for target detection via Kalman filter and Bayesian estimation

2018-08-07
2018-01-1608
Target detection is essential to the advanced driving assistance system (ADAS) and automatic driving. And the data fusion of millimeter wave radar and camera could provide more accurate and complete information of targets and enhance the environmental perception performance. In this paper, a method of vehicle and pedestrian detection based on the data fusion of millimeter wave radar and camera was proposed. The first step is the targets data acquisition. A deep learning model called Single Shot MultiBox Detector (SSD) was utilized for targets detection in consecutive video frames captured by camera and further optimized for high real-time performance and accuracy. Secondly, the parallel Kalman filter was used to track the targets detected by radar and camera respectively. Since targets information provided by the camera and radar are different, different Kalman filters were designed to achieve the tracking process.
Technical Paper

UWB Location Algorithm based on BP Neural Network

2018-08-07
2018-01-1605
In order to solve the problem that in the traditional trilateral positioning algorithm, the final positioning error is large when there is a certain error in the measured three-sided distance, an UWB positioning algorithm based on Back Propagation (BP) neural network is proposed. The algorithm utilizes the fast learning characteristic and the ability of approximating any non-linear mapping of neural network, and realizes the location of the mobile label through the TOA measurement value provided by the base station and the BP neural network. By comparing the traditional trilateral positioning algorithm, the BP neural network algorithm based on four distance inputs and the BP neural network algorithm based on four distance inputs with trilateral positioning coordinates, it can be seen that the positioning error of traditional trilateral positioning algorithm is 30 cm, and the positioning error of the positioning algorithm based on the BP neural network proposed in this paper is 10 cm.
Technical Paper

Vehicle State Estimation Based on Recurrent Neural Network and Road Constraints in Automated Driving

2018-08-07
2018-01-1613
In automated driving, the states of the target vehicle which could be used to characterize the vehicle behaviours are usually required for the host vehicle control. However, some key states of target vehicle are difficult to measure directly and accurately in all driving situations. In addition to this, it is hard to get the accurate parameters of the target vehicle to establish vehicle dynamics-based method which is commonly adopted to estimate the vehicle state. To address these problems, this paper investigated a novel methodology for estimating the states of target vehicle using the information gathered by several host vehicle sensors such as the camera, light detection and ranging (LiDAR) and the radar. A vehicle kinematic model based on Serret-Frenet equation was constructed, which could be used to interpret the target vehicle lateral motion.
Technical Paper

Research on Trajectory Management of Multitarget Trajectory Based on Modified Fast Algorithm for Data Association Algorithm

2018-08-07
2018-01-1619
With the development of autonomous vehicle technology, there is an increasing tendency toward the application of intelligent sensors in environment-perception system on autonomous vehicle to assist vehicle in intelligent decision making relevant to autonomous driving. As for environment-perception system, a good track management method serves as the foundation, while multitarget tracking and multisensor data fusion are recognized as the key. In this paper, a track management method is proposed to deal with multitarget tracking based on the target-level data of multisource environmental sensors for autonomous vehicle. The track management includes four procedures as following: track initiation; point-track association; track update; track deletion. A modified fast algorithm for data association is applied in the point-track association procedure. Afterwards Kalman filter is implemented to update the track information of target. The algorithm has get through a simulation test.
X