Towards automated substrate mapping with low-cost sidescan sonar

Transcript

1. Towards automated substrate mapping with low-cost sidescan sonar Daniel Buscombe,

   GCMRC-USGS MBES in Rivers Workshop Flagstaff, AZ, March 2015
2. None

3. 10cm transducer

4. Starboard scans Colorado River in Grand Canyon

5. Penobscot River, Maine

6. 30 miles of Marble Canyon in 2013 30 miles of

   Eastern Grand Canyon in 2014 Several reaches in Glen and Marble Canyons repeat mapped 2013 - 2015
7. None

8. w = cos (θ) = ℎ For given h, shadow

   length w increases as decreases θ = sin Assuming h constant over small window: Step 1: calculate w for each window Step 2: remove bias in w caused by sonar geometry Object height h and shadow length w are related by: w = w π/2 sin−1(c/tf) w = w dx

9. RM30 Marble Canyon Colorado River Preliminary data – do not

   cite

10. RM61 Marble Canyon Colorado River Preliminary data – do not

    cite

11. Dog River, Mobile May, Alabama. Credit: Allen Aven, Dauphin Island

    Sea Lab 200 kHz echosounder: full waveform 1st echo 2nd echo Chap 3.1-3.2, Penrose et al. (2005) “Acoustic Techniques for Seabed Classification” RoxAnn

12. ‘Roughness’ ‘Hardness’ Volume scattering coefficients K=3 means acoustic ‘classes’ Preliminary

    data – do not cite

13. Glen Canyon Substrate/Vegetation Mapping Lasers spaced 9cm

14. Glen Canyon Vegetation/substrate Mapping ‘Roughness’ ‘Hardness’ 3 acoustic classes Preliminary

    data – do not cite

15. PyHum (Beta) software Collaborations/data welcome!!! https://pypi.python.org/pypi/PyHum pip install PyHum https://github.com/dbuscombe-usgs/PyHum

    git clone https://github.com/dbuscombe-usgs/PyHum.git

16. Thanks! • Paul Grams, Ted Melis, Mike Yard: GCMRC •

    Sean Smith: U Maine • Dan Hamill, Joe Wheaton: USU • Allen Avery: Dauphin Island Sea Lab Upper Granite Gorge, Colorado River, May 2014

17. Upper Marble Canyon, May 2012

18. None