AHS Data Processing

For the exact geo-referencing of AHS data and the strip adjustment of the ALS data acquired in the framework of the case study Carnuntum, a dense network of ground control points was measured. After trying several approaches, a solid geo-referencing workflow for AHS data could be established together with LBI ArchPro partner ABT, who provided the software packages for geo-referencing and ortho-rectification as well as atmospheric correction.

A complete overview of Near-Infrared and Red Edge archaeological imaging has been published recently. The results were obtained with modified digital cameras, but they delivered the necessary insights into the specific behaviour of vegetation and soils in both spectral bands. Partner Uni Vienna, with collaborative partner BOKU, developed the AHS-toolbox ‘CropMark’. Although this toolbox currently holds many well-known but also completely new processing algorithms for AHS data, its data management concept is open, which means that all kinds of raster imagery (such as geophysical data) can be loaded and processed in the future. All CropMark functions were evaluated and a methodology developed for comparing all the images that are produced. Especially worthwhile are the use of Red Edge Inflection Point and a distribution fitting based on a Gamma curve.

First tests with an object-oriented semi-automatic approach have been undertaken for the classification of multi-temporal AHS data sets, starting with methodological questions regarding initially pixel-based algorithms.

Universität für Bodenkultur Wien