Overview of autonomous vehicle related activities

yakzephyrAI and Robotics

Nov 24, 2013 (4 years and 1 month ago)

58 views

Overview of autonomous vehicle related activities



D.Gibbins
, October 2010

Sensor Signal Processing Group, EEE dept, Adelaide Uni, October 2010

SSP Group Overview


Team of 4
-
5 researchers plus
Phd

Students
(Research Leader: Prof.
D.A.Gray
)


Specialising in Signal (& Information) Processing


Radar
(L
-
band, SAR, ISAR , phased
-
array, MIMO)


Electro
-
optical, LIDAR/LADAR, Sonar sensors etc..


GPS/INS


Target classification, recognition, 2D image and 3D scene analysis,
route planning etc


Focus on applications related to Autonomous vehicles


GPS Anti
-
jam, jammer localisation (single/multiple UAV’s)


Sensor fusion, path planning using PMHT, SLAM etc...


Terrain & scene analysis


Target recognition
(2D & 3D
)


apps in aerial surveillance



Radar sensors for autonomous vehicles
(research interest)


Detection/mapping/collision avoidance?


Sensor Signal Processing Group, EEE dept, Adelaide Uni, October 2010

Angular Separation of Sources (degrees)
Interference DOA (degrees)
Beampattern as Mobile Interference Approaches North Camp
10
20
30
40
50
60
70
0
20
40
60
80
100
120
140
160
180
Angular Separation of Sources (degrees)
Interference DOA (degrees)
Beampattern as Mobile Interference Approaches North Camp
10
20
30
40
50
60
70
0
50
100
150
Conventional and improved
interference localisation

GPS


Interference Mitigation &
Localisation for UAV
applications


Temporal, spatial and STAP processing


Adaptive beam
-
forming


Null steering


DOA estimation


Successful anti
-
jam trials held in
Woomera in presence of multiple
interference sources


Ongoing development of compact anti
-
jam hardware for aerial platforms

Principle Researcher: Matthew
Trinkle

-8000
-6000
-4000
-2000
0
2000
4000
6000
-2000
0
2000
4000
6000
8000
10000
12000






























South

East

Measured Van Location

Unprotected Receiver

Measurements

Protected Receiver

Measurements


Sensor Signal Processing Group, EEE dept, Adelaide Uni, October 2010

UAV surveillance & targeting


Electro
-
optical Seeker Target Recognition
(DSTO
sponsored)


Static land based & littoral moving targets etc


LADAR/LIDAR terrain reconstruction and
classification
(DSTO sponsored)


Stabilisation, reconstruction & scene analysis for apps such
as route planning, situation awareness etc


LADAR/LIDAR 3D target recognition
(DSTO & self funded
R&D)


ICP registration, SIFT matching, correlation based etc
(high res
and more recently low
-
resolution data)


Video based stabilisation/super
-
resolution/geo
-
location
(DSTO sponsored)

Principle Researcher: Danny
Gibbins

Sensor Signal Processing Group, EEE dept, Adelaide Uni, October 2010

“A Comparison of Terrain
Classification using Local Feature
measurements of 3
-
Dimensional
Colour Point
-
cloud Data”
D.Gibbins

IVCNZ 2009.

EO Mid
-
course Navigation,
LADAR Terrain
Analysis & Classification


Example
of EO Model Recognition for
navigation correction


Real Data

3D Terrain
reconstruction from
airborne LADAR &
optical data

Sensor Signal Processing Group, EEE dept, Adelaide Uni, October 2010

3D Sift
feature
analysis

3D Sift
feature
matching

3
D LADAR/LIDAR Target Recognition (& registration)


“3D Target Recognition Using 3
-
Dimensional
SIFT or Curvature Key
-
points and Local Spin
Descriptors”
D.Gibbins

DASP 2009.

Sensor Signal Processing Group, EEE dept, Adelaide Uni, October 2010

PMHT Path Planning for UGV’s
(
Cheung,Davey,Gray
)


Probabilistic multi
-
hypothesis tracking for
UGV path planning


Treats locales of interest as
measurements and UGV
platforms as targets


Attempts to optimise search
across multiple UGV’s


z
1

z
n

k
1
;
π
k

k
n
;
π
k

Waypoint to
platform
assignments

Waypoints

Platfor
m
States

x
0
1

x
1
1

τ
1
;
π
τ

τ
t
;
π
τ

Waypoint to
time
assignments

x
0
m

x
T
1

x
T
m

z
2

k
2
;
π
k

τ
2
;
π
τ

x
1
m



Example
of path
planning
for 4
UGV’s
based on
random
locations
of
interest