High spatial resolution data acquisition for the geosciences: kite aerial photography

Mike J. Smith, Jim H. Chandler, James Rose

Earth Surface Processes and Landforms · 2008 · 124 citations · 16 references

DOIFull text

Open access

Concepts

TL;DR

High‑resolution (<0.25 m) elevation data are essential for detailed morphometric analyses, but conventional airborne or heliborne methods are too costly for small studies, prompting the development of kite aerial photography (KAP) using a kite and consumer digital camera. KAP employs a kite‑mounted consumer camera, with ground control points surveyed at three test sites, and images processed in Leica Photogrammetry Suite to evaluate accuracy via photogrammetry. The experiments show that KAP can generate many accurately located points when image texture is adequate, though data suitability ultimately depends on assessing measurement variability and acceptable accuracy thresholds. © 2008 John Wiley & Sons, Ltd.

Abstract

Abstract This paper highlights the requirement for very high resolution (&lt;0·25 m) elevation data for quantitative and qualitative morphometric analyses. Traditional techniques for high resolution data capture (e.g. airborne, heliborne) are prohibitively expensive for small studies and therefore a kite‐based platform was developed, in conjunction with a consumer non‐metric digital camera, for data capture. The combination of kite and digital camera is more generally termed kite aerial photography (KAP). The accuracy of data derived by digital photogrammetry and imagery acquired using a kite based non‐metric camera is assessed by three experiments: one on smooth terrain, one on tor terrain and one on a glaciofluvial esker. Ground control targets were surveyed at all three sites, with the imagery subsequently processed using the Leica Photogrammetry Suite. The results demonstrate that the method can extract a high number of sampling points at high accuracy, provided that there is suitable image texture across the site. However, final judgment concerning the suitability of derived data is dependent upon an understanding of measurement variability and user quantification of acceptable accuracy. Copyright © 2008 John Wiley &amp; Sons, Ltd.

References

16