HR: 1400h
AN: OS23E-07 [Abstracts]
TI: Improved Beach Zone Segmentation From Airborne Lidar Measurements Using Intensity Measures
AU: * Starek, M J
EM: mstarek@ufl.edu
AF: Department of Civil and Coastal Engineering
University of Florida, PO Box 116130, Gainesville, FL 32611, United States
AU: Vemula, R K
EM: vraghav@ufl.edu
AF: Department of Civil and Coastal Engineering
University of Florida, PO Box 116130, Gainesville, FL 32611, United States
AU: Slatton, C
EM: slatton@ece.ufl.edu
AF: Department of Civil and Coastal Engineering
University of Florida, PO Box 116130, Gainesville, FL 32611, United States
AU: Slatton, C
EM: slatton@ece.ufl.edu
AF: Department of Electrical and Computer Engineering
University of Florida, PO Box 116130, Gainesville, FL 32611, United States
AU: Shrestha, R L
EM: rshre@ce.ufl.edu
AF: Department of Civil and Coastal Engineering
University of Florida, PO Box 116130, Gainesville, FL 32611, United States
AU: Carter, W E
EM:
AF: Department of Civil and Coastal Engineering
University of Florida, PO Box 116130, Gainesville, FL 32611, United States
AB:
In an effort to monitor beach zone stability along the St. Augustine Beach region of Florida, high-resolution
airborne laser swath mapping (ALSM) data are routinely acquired by the University of Florida's Geosensing and
Engineering Mapping (GEM) Center. ALSM, often referred to as Light Detection and Ranging (LiDAR), systems
enable sub-meter sampling of the near-shore coastal topography and the subsequent creation of digital elevation
images with rms errors of less than 10cm over minimally-vegetated surfaces, such as beaches. Currently, there
are seven collection dates spanning August 2003 to February 2007. This high spatial resolution coupled with the
multiple acquisitions through time provided several results: two separate beach nourishment efforts were
captured in the data allowing sediment spreading rate to be modeled and volume loss quantified, shoreline
change rates were estimated for temporal scales ranging from a few months to over two years at various spatial
frequencies from < 5m to > 300m, storm and seasonal wave climate induced shoreline response were
modeled, and novel approaches to morphological feature extraction and identification of localized erosional hot-
spots were developed. All previous analyses are based on range measurements; however, the ALSM system
also records the intensity (peak voltage from the APD) for each return. Intensity has traditionally been under
utilized as a feature for image classification because it does not represent true terrain radiance. We show that in
areas with minimal topographic relief, such as beaches, intensity measures have great potential for improved
beach zone segmentation. Segmentation of the beach zone is important for several factors including identification
of the wet-dry line for traditional shoreline comparison and change-detection, and removal of water points to
allow analysis of beach-only zones. Several intensity-based features are extracted from ALSM training data
collected along the St. Augustine beach and partitioned into three classes, wet beach, dry beach, and water to
detect the water line. Class-conditional probability density functions are estimated for each feature to assess
which are most informative and their separability is ranked. Results indicate significant class separation using
centroidal features, such as mean and median, suggesting robust segmentation of the beach using intensity
measures is possible. The method presented provides a novel geometric feature extraction and a systematic
feature selection procedure for high-resolution ALSM intensity data.
DE: 4262 Ocean observing systems
DE: 4294 Instruments and techniques
DE: 4546 Nearshore processes
DE: 9805 Instruments useful in three or more fields
SC: Ocean Sciences [OS]
MN: 2007 Joint Assembly