GPR Activities within the German Project MMSR Vehicle borne Mine Detection System

1
GPR Activities within the German Project
MMSR(Vehicle borne Mine Detection System)

• Third DTIF Workshop23. - 24. September 2002JCR
  Ispra, Italy

Jörg LaukemperRheinmetall Landsysteme GmbH,
Germany
2
Table of Content

• MMSR System
• GPR Sensor
• Field Trials
• Next Steps

3
Research Project under contract of the German
BWB Department KG IV 3
MMSRMobiles Minen-Such und Räumgerät(Mobile
Mine Detection and Clearing Device)

• Start of 1. program phase Sept. 1997
• End of 1. Program phase Dec. 2001
• Evaluation of the field trials 10.9. - 13.9.2001
  at Unterlüß

4
Objectives of MMSR Phase 1

• Development of an experimental demonstrator to
• determine the mine detection capabilities of a
  real time multi sensor system with real time data
  fusion
• find out the optimum sensor combination and
  configuration
• specify the sensor characteristics for an
  operational MMSR system for phase 2

Consortium of companies Rheinmetall, STN Atlas,
BGT
5
MMSR SENSOR TECHNOLOGIES

• Electro-magnetic Metal Detector (EMD)
• Multi-spectral Camera Platform (MSP)
• Ground Penetrating Radar (GPR)(All Sensors have
  a detection width of 2 m)
• In Future Verification (Explosive) Sensor

6
MMSR EXPERIMENTAL SENSOR CARRIER
GPS Antenna
MSP
GPS Antenna
Control Cabin
GPR antenna
EMD arrays
7
GPR SENSOR
Antenna multiplexer electronic
Antenna

• Type Broad band pulse echo
• Antenna elements 38
• Width 2 m

8
GPR ANTENNA WITH MINIMUM CLUTTER
Test Configuration
Resulting Signal
(Signal / Clutter Ratio gt 50 dB)
(resistively loaded)
9
GPR ANTENNA DESIGN
3 dB-Footprint18 cm at 20 cm distance
10
GPR PULSE CHARACTERISTIC
Centre Frequency 2.5 GHz
Pulse width 0.40 ns
Best compromise between ground penetration and
spatial resolution
11
GPR RAW DATA OF A METALLICAT MINE AT 23 cm DEPTH
Longitudinal View of Antenna No. 16 B- Scan
Horizontal View at 4.1 ns C-Scan
Number of Antenna
Sample Time in Depth ns
Longitudinal Sample Number
Longitudinal Sample Number
12
GPR RAW DATA OF A NONE METALLICAT MINE AT 10 cm
DEPTH
Longitudinal View of Antenna No. 16 B-Scan
Horizontal View at 4.1 ns C-Scan
Number of Antenna
Sample Time in Depth ns
Longitudinal Sample Number
Longitudinal Sample Number
13
GPR RAW DATA OF A LOW METALAP MINE AT 10 cm DEPTH
Longitudinal View of Antenna No. 16 B-Scan
Horizontal View at 4.1 ns C-Scan
Number of Antenna
Sample Time in Depth ns
Longitudinal Sample Number
Longitudinal Sample Number
14
Objectives of Field Trial

• Evaluation of overall MMSR system performance and
  evaluation of performance of each Sensor
  regarding
• Probability of Detection
• False Alarm Rate
• on realistic prepared Mine Fields

15
Preparation of Trials

• Scenario (following Mine Fields in Kosovo)
• Paved Road
• Dirt Track
• Gravel Road / Verge
• Off Road Heathland
• Threat
• 56 AP Mines (open, embedded in surface)
• 31 AT Mines (open, buried, embedded in surface)

16
Operational ScenarioOff Road Heathland
Original Terrain in Kosovo with PMR 2A Mine
barrier
Test Terrain Heathland
17
Operational Scenario Dirt Track / Gravel Road
Test Terrain Gravel Road
Original Terrain in Kosovo
Test Terrain Verge
Test Terrain Dirt Track
18
Types of Mines(Anti Tank Mines)
TMA 3
TM62 M
metallic
none metallic
TMA 2
none metallic
19
Types of Mines(Anti Personnel Mines)
PMD 6
PMA 2
none metallic
none metallic
PMR
none metallic
PMN
PMR 2A
low metal
metallic
20
ScenarioGravel Road with Mine Cells
Grass
Grass
Gravel
Closed Mine Cell
TM62 M
TMA 2
TMA 3
21
Scenario Dirt Track
Tracks
PMR 2A
PMA 2 partly covered
TMA 3 covered
22
Scenario Off Road Heathland
TMA 2
PMR 2A
PMN
PMA 2 partly covered
PMR
PMD 6
TM62 M
TMA 3
23
Trial Execution

• Online Signal Processing and Object Detection
• Position referenced (RTK GPS) Data Acquisition
• Scanned Mines 313 (174 AP / 139 AT)
• Scanned Area 7581 m2
• EnvironmentalConditions dry / wet

24
Trial ResultsProbability of Detection
AT Mines 98.5(1 buried metal free mine was
missed twice in wet heathland scenario) AP
Mines 55 (with PMA 2)90 (without PMA 2)
PMA 2 (none metallic diam. 6 cm) embedded in the
surface was not detected
25
False Alarms(high sensitivity)
EMD
GPR
Optronic
4
5
False Alarm / m2
m2 / False Alarm
7
10
20
Gravel Road
Dirt Track
Asphalt Road
Off Road Heathland
Verge
26
Next StepsGeneral

• Development of a near-operational MMSR System
• Optimisation of sensors

27
Next StepsImprovement of PoD

• Improvement of Detection Probability of none
  metallic AP mines by
• increasing spatial resolution of MSP for open
  laid mines
• triggering none detected AP mines(Quality
  assurance platform)

28
Next StepsReduction of FAR
Reduction of False Alarm Rate by use of Explosive
Detector(TNA Technology) Verification of each
Alarm before Clearing Procedure starts
29
QuestionsorComments?
30
Results of Detection(Anti Tank Mines)
TMA 3
none metallic
detected
TM62 M
detected
TMA 2
metallic
none metallic
detected (only partly in Off Road Terrain)
31
Results of Detection(Anti Personnel Mines)
PMR 2A
metallic
detected
PMR
none metallic
partly detected
32
Results of Detection(Anti Personnel Mines)
PMA 2
none metallic
not detected
PMD 6
PMN
none metallic
metallic
partly detected
detected
33
(No Transcript)
34
GPR Antenne Signal-Timing
Radar-Video-Signal Bildfrequenz 20 Hz
Kanalfrequenz 780 Hz (38 Kanäle) Pulswiederholfre
quenz 15 MHz Downsampling mit ca. 20000
Abtastungen/Kanal
Digitalisierung Abtastfrequenz eff. 400kHz 512
Werte (16 Bit) pro Kanal parallel Schnittstelle
800kBaud RS422 (differentiell)