Projects:

Ecosynth: 3D Ecological Mapping Using Computer Vision

 

The Ecosynth process generates 3D scans of vegetation from sets of digital photographs acquired at low altitude from light-weight aerial platforms or the ground (Dandois and Ellis 2010).

As illustrated at right, computer vision algorithms generate 3D "point clouds" by building geometry from matching features identified in multiple overlapping photographs. 3D point clouds are then georeferenced and used to make ecological measurements.

The project began in Spring 2009 with support from the USFS and the UMBC CUERE IGERT.

The main Ecosynth page is here.

   Ecosynth Blog
 - Overall project progres as well as weekly updates from the whole    team on progress with aircraft, cameras, and ecology.
   Subscribe to the Ecosynth Blog by Email



The first Ecosynth publication in the open-source journal Remote Sensing:

Dandois, J. P. and E. C. Ellis. 2010. Remote sensing of vegetation structure using computer vision. Remote Sensing 2(4):1157-1176. [download]

Funders & Support
USDA USFS Shield    Baltimore Ecosystem Study   

Forestry research is supported by USDA Forest Service joint venture agreement 06-JV-11242300-135. 
Undergraduate student research supported by an NSF REU from the Baltimore Ecosystem Study DEB-0423476.
Graduate student research and some equipment supported by NSF IGERT 054969, hosted by UMBC CUERE.


Contact Erle Ellis for more information.