Low-Level Image Processing for Lane Detection and Tracking

<span lang="EN-US">In this article, a new method of vehicles detecting and tracking is presented: A thresholding followed by a mathematical morphology treatment are used. The tracking phase uses the information about a vehicle. An original labeling is proposed in this article. It helps to reduce some artefacts that occur at the detection level. The main contribution of this article lies in the possibility of merging information of low level (detection) and high level (tracking). In other words, it is shown that many artefacts resulting from image processing (low level) can be detected, and eliminated thanks to the information contained in the labeling (high level). The proposed method has been tested on many video sequences and examples are given illustrating the merits of our approach.</span>

Download Full-text

Vision System Based Lane Keeping Assistance Evaluated in a Hardware-in-the-Loop Simulator

ASME 2010 10th Biennial Conference on Engineering Systems Design and Analysis, Volume 3 ◽

10.1115/esda2010-24346 ◽

2010 ◽

Cited By ~ 1

Author(s):

Fuat Cos¸kun ◽

O¨zgu¨r Tuncer ◽

Elif Karslıgil ◽

Levent Gu¨venc¸

Keyword(s):

Image Processing ◽

Real Time ◽

Lane Detection ◽

Assistance System ◽

Hardware In The Loop ◽

Vehicle Model ◽

Detection And Tracking ◽

Video Frames ◽

Assistance Systems ◽

Lane Keeping

Lane keeping assistance systems help the driver in following the lane centerline. While lane keeping assistance systems are available in some mass production vehicles, they have not found widespread use and are not as common as ESP or ACC at the moment. Lane keeping assistance systems still need further development. Previously available systems have to be continuously adapted to newer vehicle models and fully tested after this adaptation. An image processing algorithm for lane detection and tracking, a lane keeping assistance controller design and a real time hardware-in-the-loop (HiL) simulator developed for testing the designed lane keeping assistance system are therefore presented in this paper. The high fidelity, high order, realistic and nonlinear vehicle model in Carmaker HiL runs as software in a real time simulation on a dSpace compact simulator with the DS1005 and DS2210 boards. A PC is used for processing video frames coming from an in-vehicle camera pointed towards the road ahead. Lane detection and tracking computations including fitting of composite Bezier curves to curved lanes are carried out on this PC. In the present setup, the camera used is a virtual camera attached to the virtual vehicle in Carmaker and provides video frames from the Carmaker animation screen. A dSpace microautobox is available for obtaining the lane data from the PC and the Carmaker vehicle data from the dSpace compact simulator and calculating the required steering actions and sending them to the Carmaker vehicle model. The lane keeping controller is designed in the Matlab toolbox COMES using parameter space techniques. The motivation behind this approach is to develop the lane keeping assistance system as much as possible in a laboratory hardware-in-the-loop setting before time consuming, expensive and potentially dangerous road testing. Lane detection, tracking and curved lane fit results, hardware-in-the-loop simulation results of the lane keeping controller with the image processing system are are used to demonstrate the effectiveness of the proposed method.

Download Full-text

Panoramic Vision System for Autonomous Driving Vehicle

Applied Mechanics and Materials ◽

10.4028/www.scientific.net/amm.644-650.497 ◽

2014 ◽

Vol 644-650 ◽

pp. 497-501

Author(s):

Yu Bin Zhou

Keyword(s):

Image Processing ◽

High Performance ◽

Vision System ◽

Autonomous Driving ◽

Intelligent Vehicles ◽

Lane Detection ◽

Detection And Tracking ◽

Panoramic Vision ◽

Processing Module ◽

Time Image

High effective vision system is important for autonomous driving vehicles. A panoramic vision system based on FPGA+DSP with 6-camera for intelligent vehicles is presented in this paper. The system includes digital image acquisition module and high image processing module which work independently to each other. The including two C6416 DSP chips and one high-performance Virtex-4 FPGA to achieve the complex real-time image processing during autonomous driving, such as cylindrical panoramic image rebuilding, lane detection and tracking. The proposed algorithm was also optimized according to the specific characteristics of the hardware for high parallel processing in FPGA and pipelined in DSP.

Download Full-text