Huiyan Chen

A novel lane detection based on geometrical model and Gabor filter

Many people die each year in the world in single vehicle roadway departure crashes caused by driver inattention, especially on the freeway. Lane Departure Warning System (LDWS) is a useful system to avoid those accident, in which, the lane detection is a key issue. In this paper, after a brief overview of existing methods, we present a robust lane detection algorithm based on geometrical model and Gabor filter. This algorithm is based on two assumptions: the road in front of vehicle is approximately planar and marked which are often correct on the highway and freeway where most lane departure accidents happen [1]. The lane geometrical model we build in this paper contains four parameters which are starting position, lane original orientation, lane width and lane curvature. The algorithm is composed of three stages: the first stage is called off-line calibration which just runs once after the camera is mounted and fixed in the vehicle. The parameters of camera used for lane detection is accurately estimated by the 2D calibration method [2]; The second stage is called lane model parameters estimation and lane model candidates construction, the first three parameters, starting position, lane original orientation and lane width will be estimated using dominant orientation estimation [3] and local Hough transform. Then the construction of lane model candidates is implemented for the final lane model matching; the third stage is model matching. The proposed lane module matching algorithm is implemented to match the best fitted lane model. The combination of these modules can overcome the universal lane detection problems due to inaccuracies in edge detection such as shadow of tree and passengers on the road. Experimental results on real road will be presented to prove the effectiveness of the proposed lane detection algorithm.

The recognition and tracking of traffic lights based on color segmentation and CAMSHIFT for intelligent vehicles

The recognition and tracking of traffic lights for intelligent vehicles based on a vehicle-mounted camera are studied in this paper. The candidate region of the traffic light is extracted using the threshold segmentation method and the morphological operation. Then, the recognition algorithm of the traffic light based on machine learning is employed. To avoid false negatives and tracking loss, the target tracking algorithm CAMSHIFT (Continuously Adaptive Mean Shift), which uses the color histogram as the target model, is adopted. In addition to traffic signal pre-processing and the recognition method of learning, the initialization problem of the search window of CAMSHIFT algorithm is resolved. Moreover, the window setting method is used to shorten the processing time of the global HSV color space conversion. The real vehicle experiments validate the performance of the presented approach.

Road detection using support vector machine based on online learning and evaluation

Road detection is an important problem with application to driver assistance systems and autonomous, self-guided vehicles. The focus of this paper is on the problem of feature extraction and classification for front-view road detection. Specifically, we propose using Support Vector Machines (SVM) for road detection and effective approach for self-supervised online learning. The proposed road detection algorithm is capable of automatically updating the training data for online training which reduces the possibility of misclassifying road and non-road classes and improves the adaptability of the road detection algorithm. The algorithm presented here can also be seen as a novel framework for self-supervised online learning in the application of classification-based road detection algorithm on intelligent vehicle.

A GA Based Combinatorial Auction Algorithm for Multi-Robot Cooperative Hunting

How to make a trade off between the security of sys- tems and the usability of users is an important issue in network security configuration. To resolve this problem, an optimization method of security configuration based on game theory is proposed. Firstly, a security configuration model based on non-cooperative game is built which in- fers the optimal strategy of systems and users respectively by calculating their strategies and incentives. Secondly, to optimize security configuration further, it cooperatively optimizes the individual optimal strategy by cooperative game, thus eliminating the cases that individual optimal is not the overall optimal. An illustrated experiment shows that this method can coordinate the security of network systems and usability of users so that the security configu- ration of system is optimized magnificently.In order to improve the hunting efficiency of multi- robot cooperative hunting in complicated environment: multi-target and dynamic continues surrounding, a combinatorial auction model based on genetic algorithm (GACA) was presented in this paper. The model adopted genetic algorithm to solve the winner determination problem in combinatorial auction. We also compared the combinatorial auction model based task allocation method with the traditional single item auction model in solving dynamic and complex task allocation problem in multi-robot cooperation. The simulation experiments were conducted in a self- developed visible multi-robot simulation platform, OpenSim, and the results showed the whole process of hunting was very smooth, and the cost time cost by our algorithm was much shorter than the compared method.

High Speed Lane Recognition under Complex Road Conditions

To improve the speed and stability of lanes recognition under complex road conditions, a rapid detection algorithm combined with dynamic window and prior knowledge is proposed. The method of grid is used to segment initial image and define the region of interest (ROI), then all the pixels are eliminated except those on the intersection of the grid lines, thus feature pixels of the lane edge in these intersections will be detected and can be used to generate some dynamic windows with the dilatation algorithm. Then, images are processed in these dynamic windows to obtain lane edge feature. Slope and intercept of the lane can be attained by Hough transformation. To obtain the correct lane information, the information obtained in current processing cycle will be judged and filtered by the prior knowledge. Experiments in structured road showed that the speed of image processing was about 22ms/frame and the proposed algorithm could meet the real-time and stability requirements of high-speed vehicle vision navigation system.

An iterative linear quadratic regulator based trajectory tracking controller for wheeled mobile robot

We present an iterative linear quadratic regulator (ILQR) method for trajectory tracking control of a wheeled mobile robot system. The proposed scheme involves a kinematic model linearization technique, a global trajectory generation algorithm, and trajectory tracking controller design. A lattice planner, which searches over a 3D (x, y, θ) configuration space, is adopted to generate the global trajectory. The ILQR method is used to design a local trajectory tracking controller. The effectiveness of the proposed method is demonstrated in simulation and experiment with a significantly asymmetric differential drive robot. The performance of the local controller is analyzed and compared with that of the existing linear quadratic regulator (LQR) method. According to the experiments, the new controller improves the control sequences (ν, ω) iteratively and produces slightly better results. Specifically, two trajectories, ‘S’ and ‘8’ courses, are followed with sufficient accuracy using the proposed controller.

Hybrid Map-Based Navigation Method for Unmanned Ground Vehicle in Urban Scenario

To reduce the data size of metric map and map matching computational cost in unmanned ground vehicle self-driving navigation in urban scenarios, a metric-topological hybrid map navigation system is proposed in this paper. According to the different positioning accuracy requirements, urban areas are divided into strong constraint (SC) areas, such as roads with lanes, and loose constraint (LC) areas, such as intersections and open areas. As direction of the self-driving vehicle is provided by traffic lanes and global waypoints in the road network, a simple topological map is fit for the navigation in the SC areas. While in the LC areas, the navigation of the self-driving vehicle mainly relies on the positioning information. Simultaneous localization and mapping technology is used to provide a detailed metric map in the LC areas, and a window constraint Markov localization algorithm is introduced to achieve accurate position using laser scanner. Furthermore, the real-time performance of the Markov algorithm is enhanced by using a constraint window to restrict the size of the state space. By registering the metric maps into the road network, a hybrid map of the urban scenario can be constructed. Real unmanned vehicle mapping and navigation tests demonstrated the capabilities of the proposed method.

Intention-Aware Autonomous Driving Decision-Making in an Uncontrolled Intersection

Autonomous vehicles need to perform social accepted behaviors in complex urban scenarios including human-driven vehicles with uncertain intentions. This leads to many difficult decision-making problems, such as deciding a lane change maneuver and generating policies to pass through intersections. In this paper, we propose an intention-aware decision-making algorithm to solve this challenging problem in an uncontrolled intersection scenario. In order to consider uncertain intentions, we first develop a continuous hidden Markov model to predict both the high-level motion intention (e.g., turn right, turn left, and go straight) and the low level interaction intentions (e.g., yield status for related vehicles). Then a partially observable Markov decision process (POMDP) is built to model the general decision-making framework. Due to the difficulty in solving POMDP, we use proper assumptions and approximations to simplify this problem. A human-like policy generation mechanism is used to generate the possible candidates. Human-driven vehicles’ future motion model is proposed to be applied in state transition process and the intention is updated during each prediction time step. The reward function, which considers the driving safety, traffic laws, time efficiency, and so forth, is designed to calculate the optimal policy. Finally, our method is evaluated in simulation with PreScan software and a driving simulator. The experiments show that our method could lead autonomous vehicle to pass through uncontrolled intersections safely and efficiently.

Optimal design and simulation evaluation of economical gear-shifting schedule for AMT in pure electronic bus

50 “zero emission” pure electrical buses BK6122EV have been running in Beijing for the whole year. As one of the core systems of the bus, the automatic mechanical transmission (AMT) has the enormous influence on its economical performance. In order to further optimize the economic gear-shifting schedule for BK6122EV, this paper uses three design methods of the conventional engine vehicle for reference, where the fuel consumption figure of the engine is substituted for the efficiency map of the driving motor. The simulation model of BK6122EV based on AVL Cruise simulation software is built to compare and optimize the designed schedules. Moreover, the best schedule is gained. The result of the simulation indicates the economical performance of the bus using the optimized schedule is improved by 1.54 percent compared with the existing schedule.

ICP stereo visual odometry for wheeled vehicles based on a 1DOF motion prior

In this paper, we propose a novel, efficient stereo visual-odometry algorithm for ground vehicles moving in outdoor environments. To avoid the drawbacks of computationally-expensive outlier-removal steps based on random-sample schemes, we use a single-degree-of-freedom kinematic model of the vehicle to initialize an Iterative Closest Point (ICP) algorithm that is utilized to select high-quality inliers. The motion is then computed incrementally from the inliers using a standard linear 3D-to-2D pose-estimation method without any additional batch optimization. The performance of the approach is evaluated against state-of-the-art methods on both synthetic data and publicly-available datasets (e.g., KITTI and Devon Island) collected over several kilometers in both urban environments and challenging off-road terrains. Experiments show that the our algorithm outperforms state-of-the-art approaches in accuracy, runtime, and ease of implementation.