Under Water Systems. Sidescan SAS image mapping for Automatic Detection and Classification

Transcription

1 0 Sidescan SAS image mapping for Automatic Detection and Classification

2 Plan Sidescan SAS image mapping for ADC Sidescan Synthetic Aperture Sonar Images Interest of SAS for automatic detection and classification Example of automatic detection on waterfall SAS Sonar data mapping Interest of mapping Examples SAS sonar data mapping and change detection Mapping large area Adding automatic detections to sonar mosaic Conclusions 1

3 Sidescan Synthetic Aperture Sonar Principle: Coherent integration of several pings along array displacement Increased resolution compared to physical array aperture Examples Resolution gain of 10 2

4 Sidescan Synthetic Aperture Sonar Other example with towed sidescan SAS Wedding Cake Ø96cm Manta Ø100cm 3

5 Sidescan Synthetic Aperture Sonar : Swansea Wreck 4

6 Sidescan Synthetic Aperture Sonar High resolution increases detection and classification performances: For an operator For an automated system SAS gives the opportunity to increase autonomy of AUV Automatic detection of small objects Automatic classification of known targets Sea bed classification Sonar performances evaluation Open the access to superior level of autonomy (re-planification) TUS has developed Automatic Detection and Classification for AUV and surveillance systems: Automatic detection and class can be added on the sonar waterfall See example on next slide 5

7 Waterfall representation of Sidescan SAS data 6

8 SAS Sonar data mapping Sidescan SAS image mapping for ADC Waterfall is a good way to present data in real time to an operator For AUVs, a large amount of data is available at the end of the mission Interest of having a synthetic view of the complete mission Sonar data are mapped directly using vehicle position A new sonar line is added to the mosaic, new data replace older data 7

9 SAS Sonar data mapping Transparency can be useful to compare data It can be used for detection of new object in a known area 8

10 SAS Sonar data mapping Transparency can be useful to compare data It can be used for detection of new object in a known area 9

11 SAS Sonar data mapping: Transparency and change detection 10

12 SAS Sonar data mapping: Transparency and change detection 11

13 SAS Sonar data mapping: Transparency and change detection 12

14 SAS Sonar data mapping: Example of a 1Nm² area mosaic Sidescan SAS image mapping for ADC 13

15 SAS Sonar data mapping: Automatic detections are visible in one image Vehicle trajectory can be also checked The use of existing visualization tools allows easy superposition and zoom. Sidescan SAS image mapping for ADC 14

16 Zoom on interesting features 15

17 Conclusions Sidescan SAS image mapping for ADC The SAS processing gives high resolution data The quality of these data increases the performances of the automatic processing Automatic processing of real time SAS images open access to high level of autonomy for AUVs The mapping of sonar data gives a full overview of the area Automatic detections display on sonar mosaic is a powerful tool for post mission analysis Mosaic can be used also for change detection. 16

18 Meeting the Underwater Challenge Thank you for your attention 17