Title :
Assessing woody vegetation cover dynamics in the Kruger National Park, South Africa: Linking historical aerial photographs and spot imagery
Author :
Christopher Munyati;Ndidzulafhi I. Sinthumule
Author_Institution :
Department of Geography and Environmental Science, North-West University, Private Bag X2046, Mmabatho, 2735, South Africa
fDate :
7/1/2015 12:00:00 AM
Abstract :
The coexistence of woody vegetation and grass characterises savannas. High spatial resolution digital satellite images and appropriate image processing algorithms can facilitate monitoring of the woody cover. Historical aerial photographs enable an extension of high spatial resolution analysis of the woody cover to periods before satellite imagery. This study tested a methodology that used historical (1998, 2012) SPOT-4 HRVIR gray scale (“red”) band (0.6-0.68um, 10m resolution) and SPOT-5 HRG panchromatic band (0.48-0.71 urn, 5m resolution) images in combination with panchromatic (0.4-0.7um) aerial photographs (1940, 1968, 1977; 0.44-1.35m resolution) in monitoring woody cover at sites in the Kruger National Park, South Africa. Mean Euclidean Distance texture analysis in 3×3 moving windows enhanced woody cover. The woody cover on the respective multitemporal texture images was extracted through K-means clustering, with high (>85%) indicative classification accuracy. Boolean GIS overlay analysis of the resulting woody cover thematic layers enabled woody cover change detection.
Keywords :
"Vegetation mapping","Spatial resolution","Satellites","Remote sensing","Euclidean distance"
Conference_Titel :
Geoscience and Remote Sensing Symposium (IGARSS), 2015 IEEE International
Electronic_ISBN :
2153-7003
DOI :
10.1109/IGARSS.2015.7326762