Vision-Based Autonomous Vehicle Driving Control System

In recent years, extensive research has been carried out on autonomous vehicle system. A completely autonomous vehicle is one in which a computer performs all the tasks that the human driver normally would. However, this study only focuses on driving control system that based on vision sensor. Th...

Full description

Saved in:
Bibliographic Details
Main Author: Isa, Khalid
Format: Thesis
Language:English
English
Published: 2005
Subjects:
Online Access:http://psasir.upm.edu.my/id/eprint/6052/1/FK_2005_45.pdf
http://psasir.upm.edu.my/id/eprint/6052/
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In recent years, extensive research has been carried out on autonomous vehicle system. A completely autonomous vehicle is one in which a computer performs all the tasks that the human driver normally would. However, this study only focuses on driving control system that based on vision sensor. Therefore, this study presents a simulation system with Graphical User Interface (GUI) to simulate and analyse the driving control for autonomous vehicle that based on video taken fiom the vehicle during driving on highway, by using MATLAB programming. The GUI gives easy access to analyse video, image and vehicle dynamics. Once the GUI application for simulation is launched, user can enter input parameters value (number of frames, canny edge detection value, vehicle speed, and braking time) in text control to simulate and analyse video images and vehicle driving control. In this study, there are four subsystems in the system development process. The first subsystem is sensor. This study was used a single Grandvision Mini Digital Video as sensor. This video camera provides the information of Selangor's highway environment by recording highway scene in front of the vehicle during driving Then, the recorded video is process in second subsystem or named as imageprocessing subsystem. In this subsystem, image-capturing techniques capture the video images frame by frame. After that, lane detection process extracts the information about vehicle position with respect to the highway lane. The results are angle between the road tangent and orientation of the vehicle at some look-ahead distance. Driving controller in the controller subsystem that is the third subsystem used the resulted angle from lane detection process along with vehicle dynamics parameters to determine the vehicledriving angle and vehicle dynamics performance. In this study, designing a vehicle controller requires a model of vehicle's behaviour whether dynamics or kinematics. Therefore, in vehicle subsystem that is the fourth subsystem, this study used vehicle's dynamics behaviour as the vehicle model. The model has six degrees of fieedom (DOF) and several factors such as the vehicle weight, centre of gravity, and cornering stifkess were taken into account of dynamics modelling. The important contribution of this study is the development of vehicle lane detection and tracking algorithm based on colour cue segmentation, Canny edge detection and Hough transform. The algorithm gave good result in detecting straight and smooth curvature lane on highway even when the lane was afTected by shadow. In this study, all the methods have been tested on video data and the experimental results have demonstrated a fast and robust system.