This Project was supported by Italian CNR through the
`Progetto Finalizzato Trasporti 2'.
The Italian PROMETHEUS Research Team working on PRO-ART and PRO-CHIP
Projects presents MOB-LAB, a MOBile LABoratory suited to develop and test real
time intelligent systems, monitoring the trajectory of a vehicle running on
extraurban roads, by means of computer vision systems.
A few images of the MOB-LAB vehicle
are available thanks to the courtesy of
Jose Pettiti (CSTV - Torino)
Obstacle and Road Detection. Road following and obstacle detection in
front of the vehicle are performed by means of a real time vision system
integrating techniques based on visual saliency with depth cues obtained by
stereometry and a-priori knowledge of the road environment. The environment
around the vehicle and the dangerous situations are summarized by means of nine
leds of different colors. This function has been realized by the two Pro-Art
research units of Turin: Istituto Elettrotecnico Nazionale "G.Ferraris" and
Centro Studi per la Televisione of the CNR.
Lane Tracking Control. Lane tracking is performed through the
analysis of data coming from a TV camera and interpreted by means of a special
purpouse massively parallel architecture for real-time image analysis, PAPRICA.
The result of the analysis is presented to the driver through a set of five leds
of different colors. More precisely the leds provide a visual warning to the
driver when the car approaches the lane borders, becoming more evident when more
dangerous positions are taken. This function has been realized by the Pro-Art
research unit of the University of Parma, using the low-cost massively parallel
board developed by the Pro-Chip research unit of the Politecnico of Turin and of
the University of Parma.
Cruise Control. This function is to estimate the security distance
from the in-front vehicle by means of data acquired through a TV camera. The
three leds in the control panel tell the driver the current position of the in-
front vehicle: no light on for no vehicle; green for a vehicle at a security
distance; yellow for warning; red for danger. This function has been realized by
the Pro-Art research unit of the University of Genoa.
Lateral Control. The lateral position of the vehicle with respect to
a reference line (guard rail or continuous lateral emergency line), is detected
by means of a millimeter-wave radar mounted on the dashboard. The detection of
the reference line is made possible by positioning electromagnetic reflectors
along the reference line. The data coming from the radar, fused with the
information on the speed of the vehicle, are interpreted in order to assist the
driver in preventing the vehicle from going off the road, and presented to the
driver by means of three leds of different colors. This function has been
realized by the Pro-Chip research unit Centro Studi Propagazione ed Antenne of
Overtaking Control. This funtions is to alert the driver during an
overtaking manoeuvre if another vehicle is coming from the back. The detection
of the incoming vehicle is performed by processing the images acquired from a TV
camera looking outside the vehicle from the back glass. The leds in the control
panel tell the driver the current position of the overtaking vehicle: green for
no vehicles on the back; yellow if a vehicle is present on the back; red if a
vehicle has initiated an overtaking manoeuvre. This function has been realized
by the Pro-Art research unit of the University of Genoa.
Car attitude Control. Different low-cost sensors inside the vehicle
measure the longitudinal and trasversal components of the acceleration. Data
coming from the sensors are interpreted and presented to the driver by means of
visual information and alarms (to be specified). The sensors have been designed
and manufactured by the Pro- Chip research units of the University of Brescia
and of the University of Modena .
The system shows the results of different approaches of computer vision
techniques to alert the driver in dangerous situations. The MOB-LAB vehicle has
been demonstrated at the BMM ( Board Member Meeting) in Paris, October 1994.
Prof. Giovanni Adorni,
Department of Computer Engineering - University of Parma,
Viale delle Scienze - 43100 Parma - ITALY
PRO-ART Research Units:
Department of Communication, Computer and System Science - University of Genova, Prof. Aldo Grattarola and Dr. Massimo Tistarelli