BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:VEdge_Detector: A tool for automated coastal vegetation edge detec
 tion using very deep convolutional neural networks - Martin Rogers Univers
 ity of Cambridge
DTSTART:20200501T100000Z
DTEND:20200501T113000Z
UID:TALK142066@talks.cam.ac.uk
CONTACT:Jonathan Rosser
DESCRIPTION:Coastal communities\, land covers and inter-tidal habitats in 
 East Anglia are particularly vulnerable to both erosion and flooding. This
  vulnerability is likely to increase with sea level rise and greater storm
 iness in the near future. The determination of shoreline position\, and it
 s landward migration\, is therefore imperative for future coastal risk ada
 ptation. This project is developing an automated tool\, VEdge_Detector\, t
 o extract the coastal vegetation line from high spatial resolution (Planet
 's 3-5m) imagery\, training a very deep convolutional neural network (VGGN
 et-16) to predict sequential vegetation line locations. Red\, Green and Ne
 ar-Infrared (RG-NIR) was found to be the optimum image band combination du
 ring neural network training and validation. VEdge_Detector outputs were c
 ompared with in-situ vegetation line positional measurements to ascertain 
 a mean distance error of <6m (2 image pixels) and up to 86% pixel classifi
 cation accuracy. Extracting vegetation lines from Planet imagery of the ra
 pidly retreating cliffed coast at Covehithe\, Suffolk\, has identified a l
 andward retreat rate between 2010-2020 of >4ma-1. Vegetation line change o
 utputs derived from this tool are produced far quicker than other non-auto
 mated methods\, and have the potential to inform coastal risk management d
 ecisions in East Anglia and other global locations.
LOCATION:https://ukri.zoom.us/j/92946956029
END:VEVENT
END:VCALENDAR