Onboard Evolutionary Risk Recognition System for
Automobiles Toward the Risk Map System
IEEE Transactions on Industrial Electronics, 54 (2):
878--886 (April 2007)DOI: 10.1109/TIE.2007.891654
Abstract
To achieve a system that improves the safety and
comfort of the vehicle driving, a recognition system
equivalent to the human recognition ability should be
developed. However, the vehicle environment is
complicated and involves situations so diverse that a
uniform recognition processing approach cannot function
sufficiently. For a solution to this problem, we have
been studying a comprehensive risk recognition system,
which we call the risk map system, with learning
capability. As part of this paper, a system has been
developed that autonomously obtains the image
recognition processing. This paper presents a system as
an example that automatically learns through genetic
programming to obtain the image processing of
pedestrians and vehicles taken by an onboard camera
system
