Object Based Image Analysis Technique for Distinguishing Tropical Native Forest from Plantations: A Landcover Classification Analysis of Patches Surrounding the Serra do Brigadeiro State Park, Minas Gerais, Brazil

dc.contributor.advisor

Swenson, Jennifer J

dc.contributor.author

Meehan, Kelly

dc.date.accessioned

2016-04-29T20:17:19Z

dc.date.available

2017-04-29T04:30:05Z

dc.date.issued

2016-04-29

dc.department

Nicholas School of the Environment and Earth Sciences

dc.description.abstract

The Brazilian Atlantic forest is regarded as one of the hottest biodiversity hotspots in the world. Nonetheless, this area has experienced massive deforestation and significant fragmentation as a result of encroaching pasture, coffee, and eucalyptus plantations. Satellite remote sensing is one of the primary methods of monitoring forest health and extent; yet, distinguishing native forest from plantations, such as eucalyptus, remains a significant challenge in most of the tropics. This study explores the use of moderately priced, high-resolution satellite imagery and identifies the algorithms and training parameters that successfully separate eucalyptus plantations from Atlantic forest patches for future conservation prioritization.

dc.identifier.uri

https://hdl.handle.net/10161/11941

dc.subject

object based image analysis

dc.subject

RapidEye

dc.subject

Brazilian Atlantic Forest

dc.title

Object Based Image Analysis Technique for Distinguishing Tropical Native Forest from Plantations: A Landcover Classification Analysis of Patches Surrounding the Serra do Brigadeiro State Park, Minas Gerais, Brazil

dc.type

Master's project

duke.embargo.months

12

Files

Original bundle

Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
MeehanMP_FINAL.pdf
Size:
703.92 KB
Format:
Adobe Portable Document Format