Quantifying spatial heterogeneity of Coniferous trees in ATM, CASI and Eagle airborne images

Abstract

Spatial heterogeneity of airborne remote sensing images is critical for surface character delineation. The purpose of this paper is to quantify and evaluate the spatial variability and characteristic scales of Coniferous trees from multi-sensor airborne images by applying variogram modelling. The Airborne Thematic Mapper (ATM), Compact Airborne Spectrographic Imager (CASI-2), Specim AISA Eagle airborne images at Harwood, Northumberland, UK, were utilized, with spatial resolutions of 9m, 7.2m and 2.5m respectively. We demonstrate that variogram properties provide a robust assessment of the differences in spatial variability and characteristic scale between multi-sensor airborne datasets. Spatial variability of Coniferous trees in ATM airborne imagery is consistently larger than CASI airborne imagery in blue, green, red and infrared bands. The spatial variability of Eagle airborne images is strongest in red and near infrared bands but weakest in the blue band. For the blue, green, red and near infrared bands utilized, results indicate that the total within-scene variation of multi-sensor airborne images increases with wavelength. Moreover, the mean characteristic length scale consistently decreases with the nominal spatial resolution and spectral bands. It is recommended that applications of one type of tree development observations could take advantage of Eagle images in the near infrared band to gain more within-species information of spatial structure and its variability. Other applications like mapping tree species might exploit ATM images to obtain more information about spatial structure and its variability between different tree species.

DOI: 10.1109/ICSDM.2011.5969031

Extracted Key Phrases

6 Figures and Tables

Cite this paper

@article{Qiu2011QuantifyingSH, title={Quantifying spatial heterogeneity of Coniferous trees in ATM, CASI and Eagle airborne images}, author={Bingwen Qiu and Canying Zeng and Rong Long and Chongcheng Chen and Xiaoyang Tu}, journal={Proceedings 2011 IEEE International Conference on Spatial Data Mining and Geographical Knowledge Services}, year={2011}, pages={198-203} }