Graphics Reference
In-Depth Information
Chapter 10
Vision-Based Lane Analysis: Exploration
of Issues and Approaches for Embedded
Realization
Ravi Kumar Satzoda and Mohan M. Trivedi
Abstract
Vision-based lane analysis has been investigated to different degrees of
completeness. While most studies propose novel lane detection and tracking meth-
ods, there is some research on estimating lane-based contextual information using
properties and positions of lanes. According to a recent survey of lane estimation
in [
7
], there are still open challenges in terms of reliably detecting lanes in varying
road conditions. Lane feature extraction is one of the key computational steps in lane
analysis systems. In this paper, we propose a lane feature extraction method, which
enables different configurations of embedded solutions that address both accuracy
and embedded systems' constraints. The proposed lane feature extraction process is
evaluated in detail using real-world lane data to explore its effectiveness for embed-
ded realization and adaptability to varying contextual information such as lane types
and environmental conditions. Accuracy of more than 90% is obtained during the
evaluation of the proposed method using real-world driving data.
10.1 Introduction
Intelligent driver assistance systems (IDAS) are increasingly becoming a part of mod-
ern automobiles. Reliable and trustworthy driver assistance systems require accurate
and efficient means for capturing states of vehicle surroundings, vehicle dynamics
as well as state of the driver in a holistic manner [
23
].
Trivedi et al. [
23
] describe a Looking-in Looking-out (LiLo) framework for
computer-vision-based active vehicle safety systems, which aim at bringing together
the three components of overall IDAS, which are the environment, the vehicle, and the
driver. It is shown that sensing and contextualizing all the three components together
is critical and efficient in an IDAS. The role of vision sensors, complemented by
B
R.K. Satzoda (
M.M. Trivedi
Laboratory of Intelligent and Safe Automobiles, University of California San Diego,
La Jolla, CA 92093, USA
e-mail: rsatzoda@eng.ucsd.edu
M.M. Trivedi
e-mail: mtrivedi@ucsd.edu
)
ยท
