Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Schedule
(tenta:ve)
Date
1/7
1/14
1/21
Lecture
Assignment
1/28
assignment 1 due *
2/4
assignment 2 due *
2/11
2/18
2/25
3/4
3/11
8. Vehicle demo
*
changed
02/04/2013
|
ME302
Advanced
Driver
Assistance
|
Jan
Becker
|
Stanford
University
3/2/13
Assignment
1
out
of
3
assignments
must
to
be
completed
"Drive
or
be
driven?
Discuss
the
fun
of
driving
versus
the
convenience
of
driver
assistance.
one-page
essay,
due
1/28
at
2:15pm.
Details:
me302.stanford.edu
please
submit
by
email
to
jbecker@ai.stanford.edu
and
beiker@stanford.edu
02/04/2013
|
ME302
Advanced
Driver
Assistance
|
Jan
Becker
|
Stanford
University
3/2/13
Overview
Lidar
Measurement
Principle
Sensor
Variants
Lidar
Light
detecUng
and
ranging
Lidar
Principle
Distance
Measurement:
Ume
of
ight
Angular
Measurement
scanning
lidar
laser
pulse
is
reected
by
a
rotaUng
mirror
distance
measurements
at
specied
angle
intervals
non
scanning
lidar
mulU-beam
mulU-receiver
none
3/2/13
Lidar
Measurement
principle
of
scanning
lidar
source:
h_p://
upload.wikimedia.org/
wikipedia/commons/c/c0/
LIDAR-scanned-SICK-LMS-
animaUon.gif
Lidar
Sensor
Variants
ModulaUon:
pracUcally
all
lidar
sensors
use
pulse
modulaUon
Angular
Measurement
Fixed
mulU-beam
sensors
Hella
Conti
moving
opUcs
Conti
Omron
Velocity:
EsUmaUon
02/04/2013
|
ME302
Advanced
Driver
Assistance
|
Jan
Becker
|
Stanford
University
3/2/13
Lidar
ApplicaUon
Example:
Volvo
City
Safety
source: h_p://www.youtube.com/watch?v=07JUGmiEcGQ
Lidar
Sensor Example:
Lidar sensor A.D.C. AIS100
Wavelength
Eye
safety
emi_ed
power
percepUon
range
beams
range
size
velocity
accuracy
850nm
class
1
(IEC825)
12W
peak
5mW
average
11
(Azimut/hor.)
3
(ElevaUon/vert.)
sweep
5
1
to
180m
180x89x60mm
~1km/h
source: A.D.C.
10
3/2/13
Lidar
Sensor Example:
Lidar sensor Denso
percepUon
range
8
(Azimuth
/hor.)
2,2
(ElevaUon/vert.)
2D-scanning
series
sensor
in
some
Toyota
models
in
Japan
source:
Denso
11
Lidar
Sensor Example:
Lidar sensor A.D.C. AIS200
Wavelength
905nm
Eye
safety
class
1
(IEC825)
emi_ed
power
40W
peak
3,5mW
average
percepUon
range
15
(Azimut/hor.)
6,5
(ElevaUon/vert.)
scan
beams
15-30
angular
resoluUon
0.01
degrees
range
1
to
180m
size
180x89x60mm
velocity
accuracy
~1km/h
02/04/2013
|
ME302
Advanced
Driver
Assistance
|
Jan
Becker
|
Stanford
University
source: A.D.C.
12
3/2/13
Lidar
Sensor Example:
Lidar sensor Ibeo Alasca
source: Ibeo
13
Lidar
Sensor Example:
Lidar sensor Ibeo Lux
source: Ibeo
14
3/2/13
Lidar
source:
h_p://www.youtube.com/watch?v=IK29M22teJE
02/04/2013
|
ME302
Advanced
Driver
Assistance
|
Jan
Becker
|
Stanford
University
15
Lidar
MulU-Beam
Lidar
Development
16
3/2/13
Lidar
h_p://www.youtube.com/watch?v=uoiJeIb0wBA
02/04/2013
|
ME302
Advanced
Driver
Assistance
|
Jan
Becker
|
Stanford
University
17
Lidar
h_p://www.youtube.com/watch?v=uoiJeIb0wBA
02/04/2013
|
ME302
Advanced
Driver
Assistance
|
Jan
Becker
|
Stanford
University
18
3/2/13
Lidar
Velodyne
Lidar
HDL-64E
S2
64
xed
mounted
laser
laser
head
is
rotaUng
horizontal
eld
of
view:
360
verUcal
eld
of
view:
26.8
scan
rate
5-15Hz
Data
sheet:
h_p://velodynelidar.com/lidar/products/
brochure/HDL-64E
%20S2%20datasheet_2010_lowres.pdf
source:
Velodyne
02/04/2013
|
ME302
Advanced
Driver
Assistance
|
Jan
Becker
|
Stanford
University
19
source: Velodyne
Lidar
20
10
3/2/13
Lidar
21
Lidar
Why
has
Lidar
not
been
successful
(so
far)?
22
11
3/2/13
Computer
Vision
ApplicaUon
Areas
of
AutomoUve
Cameras
Rear
Camera
Parking
aid
Collision
miUgaUon/
avoidance
Front
Camera
Longitudinal
control
Lane
keeping
Collision
miUgaUon
Collision
avoidance
Night
vision
Headlamp
Control
Interior
Camera
Passenger
detecUon
Alertness
monitoring
Thes
protecUon
Surround
Camera
Parking
aid
Maneuvering
aid
23
Computer
Vision
Camera
Types
Mono
Camera
Bosch
ConUnental
Bosch
ConUnental
Stereo Camera
24
12
3/2/13
Computer
Vision
Components
lens
imager
ECU
source: Bosch
25
Computer
Vision
Imager
CCD
CMOS
photo
diode
row
adressing
select transistor
26
13
3/2/13
Computer
Vision
Imager
Types
(CMOS
/
CCD)
linear
CCD
sensor
27
Computer
Vision
Night
Vision
28
14
3/2/13
Computer
Vision
NIR
Night
Vision
source:
h_p://www.youtube.com/watch?v=DojthARCO6k
02/04/2013
|
ME302
Advanced
Driver
Assistance
|
Jan
Becker
|
Stanford
University
29
Computer
Vision
FIR
Night
Vision
source:
h_p://www.youtube.com/watch?v=-3uaTyNWcBI
02/04/2013
|
ME302
Advanced
Driver
Assistance
|
Jan
Becker
|
Stanford
University
30
15
3/2/13
Computer
Vision
Lane
Departure
Warning
source:
h_p://www.youtube.com/watch?v=GSzEi8dlFIc
02/04/2013
|
ME302
Advanced
Driver
Assistance
|
Jan
Becker
|
Stanford
University
31
Computer
Vision
Lane
Keeping
Support
source:
h_p://www.youtube.com/watch?v=GdtVe1Pnayg
02/04/2013
|
ME302
Advanced
Driver
Assistance
|
Jan
Becker
|
Stanford
University
32
16
3/2/13
Computer
Vision
Road
Sign
RecogniUon
source:
h_p://www.youtube.com/watch?v=ZaHuy5WhKAE
02/04/2013
|
ME302
Advanced
Driver
Assistance
|
Jan
Becker
|
Stanford
University
33
Computer
Vision
Rear
Camera
Parking
Aid
34
17
3/2/13
Computer
Vision
Displacement
A
displacement
vector
connects
the
projecUons
of
a
3D
point
in
the
world
in
two
images
Displacement
in
image
sequences
is
also
called
2D
moUon
Displacement
in
stereo
image
pairs
is
called
disparity
example:
image sequence
image 1
inter-frame
motion
image 2
displacement
field
35
Computer
Vision
Stereo
Vision
TriangulaUon:
Requires
recUcaUon!
base
width
les
image
right image
36
18
3/2/13
Computer
Vision
3D
reconstrucUon
of
a
trac
scene
captured
by
a
stereo
camera
system
source:
h_p://www.6d-vision.com/home/stereovision
02/04/2013
|
ME302
Advanced
Driver
Assistance
|
Jan
Becker
|
Stanford
University
37
Computer
Vision
OpUcal
Flow
Temporal
displacements
of
image
points
in
an
image
sequence
source:
h_p://www.6d-vision.com/home/opUscheruss
02/04/2013
|
ME302
Advanced
Driver
Assistance
|
Jan
Becker
|
Stanford
University
39
19
3/2/13
Computer
Vision
OpUcal
Flow
Temporal
displacements
of
image
points
in
an
image
sequence
source:
h_p://www.6d-vision.com/home/opUscheruss
02/04/2013
|
ME302
Advanced
Driver
Assistance
|
Jan
Becker
|
Stanford
University
40
Computer
Vision
Example
ApplicaUon:
AutomaUc
evasive
maneuver
source:
h_p://www.6d-vision.com/home/bedeutung
02/04/2013
|
ME302
Advanced
Driver
Assistance
|
Jan
Becker
|
Stanford
University
41
20
3/2/13
Computer
Vision
Example
Pedestrian
DetecUon
opUcal ow
stereo disparity
opUcal
ow
+
stereo
disparity
42
Summary
Lidar
Measurement
Principle
Sensor
Variants
46
21
3/2/13
Outlook
Next
lecture
Monday,
2/11/2013,
2:15pm
Bryant
Walker
Smith:
Legal
Aspects
of
Vehicle
AutomaUon
Seminar
Discussion:
Needs
and
Challenges
of
Driving
AutomaUon
47
22