Evaluation of the capability of quickbird data for automatic delineation of individual tree crowns in sparse deciduous forests. Case study in the Zagros region of Iran

This work represents first attempts at delineating individual tree crowns in Zagros area

using QuickBird data. The Zagros forest zone is the second important forest region of

Iran with high ecological and socio-economic values and provides services to a

significant parts of Iran¿s human population. These forests are experiencing rapid

degradation due to environmental and human impact at local scale. To efficiently manage

these forests for conservation and other ecological purposes forest managers demand to

have detailed information on forest composition and structure. Because of the wide extent

as well as spatial structure of these forests, extracting the required information by means

of traditional field based methods is time-consuming and costly, whilst application of

low-medium resolution satellite imagery will not yield accurate and reliable results. It is,

therefore, critically important to investigate the new advanced spaceborne remote sensing

technologies that permit to extract detail information. Very high spatial resolution satellite

imagery can supply forest management planning with individual tree based information,

which was unfeasible to accomplish with the same efficiency and precision with

moderate-resolution satellite imagery. Application to the Zagros forests face to the unique

technical challenges including high variability in crown shape and size, strong effects of

the background reflectance, and varying age and health condition.

The capabilities of QuickBird imagery examined with the data acquired over the Oak

forest site on the South West of Iran. Different image fusion techniques were applied to

exploit optimal spatial and spectral information of this imagery. The core technical

objective was selecting a fusion method with high fidelity of spectral and spatial

information. Quality assessment of fused images showed that the PCA method performs

better than the others in this study.