Object-Based Image Analysis of Salt Marsh

Submitted by Anonymous (not verified) on

Timucuan Ecological and Historic Preserve (TIMU), Jacksonville, FL, contains thousands of acres of healthy and productive salt marsh. These salt marshes provide crucial habitat to a wide variety of organisms but are threatened by a variety of influences such as sea-level rise and upland development. The National Park Service Water Resources Division is helping TIMU park management better understand landscape level changes in the marsh. Color infrared orthoimagery collected by the U.S. Geological Survey in 2012 at a 0.3-m spatial resolution, lidar data, and Trimble eCognition software are being used in an object-based image analysis technique to classify and quantify the two dominant genera (Juncus and Spartina) of salt marsh grasses found within the preserve. This technique groups neighboring homogeneous pixels and relates one group to another, enabling the analyst to accurately classify a landscape. The park will have a detailed map of the salt marsh upon completion of this project and be able to identify changes in the marsh over time.

A 3D view of the salt marsh in TIMU using data produced in this project overlaid on top of the color infrared orthoimagery and lidar elevation data. The image shows Juncus (dark green), Spartina (light green), trees (bright green), and water (blue) at a 1:1,000 scale.

A 3D view of the salt marsh in TIMU using data produced in this project overlaid on top of the color infrared orthoimagery and lidar elevation data. The image shows Juncus (dark green), Spartina (light green), trees (bright green), and water (blue) at a 1:1,000 scale.

Sensor
Platform
Author Name
Jeremy Cantor
Author Email
Jeremy_Cantor@contractor.nps.gov