In this thesis automatic classification strategies were developed on the basis of an object-oriented image analysis approach. Main goal was the support of monitoring concepts for wetlands, prescribed by recent guedelines (e.g. EU-WFD = water framework directive, FFH). Predominantly the application of spatial very high resolution IKONOS-satellite data was verified. Furthermore hyperspectral data and a time series of ortho photographs were analysed for the assessment of landscape changes. Traditional image anaylsis approaches were not able to classify spatial very high resolution remote sensing data. Thus an object-orient approach, based on rule sets and expert knowledge was used. Information could be extracted from remote sensing data and tools could be developed to support monitoring tasks of local authorities.
«
In this thesis automatic classification strategies were developed on the basis of an object-oriented image analysis approach. Main goal was the support of monitoring concepts for wetlands, prescribed by recent guedelines (e.g. EU-WFD = water framework directive, FFH). Predominantly the application of spatial very high resolution IKONOS-satellite data was verified. Furthermore hyperspectral data and a time series of ortho photographs were analysed for the assessment of landscape changes. Traditio...
»