Eine Abfragesprache für die Geometrie von Rasterelementen für die rasterorientierte kartographische Mustererkennung und Datenanalyse

This dissertation explores innovative raster-based processing techniques for topographic maps, focusing on structuring, managing, and analyzing raster map contents. It intersects geo-information systems, cartography, and pattern recognition, addressing data acquisition, pattern recognition, and data analysis in raster-based systems. Key investigations include the segmentation and computation of raster element features and the development of a geometry query language. The segmentation process isolates coherent areas in binary raster images using a run-length encoded data structure, facilitating efficient management and access. Geometrical features such as area, perimeter, and center of gravity coordinates are computed for each raster element and stored in a multidimensional feature vector. The query language, akin to SQL but enhanced with new elements, allows intuitive classification and analysis of single color layers in Swiss topographic maps. It supports re-use of previous query results, visual display of outcomes, and complex queries combining multiple single queries. This tool proves effective for pattern recognition and data analysis, enabling users to formulate queries quickly. The prototype SoftwareRaQueL (RasterQueryLanguage) implements these features, demonstrating that structured raster data enables direct pattern recognition and analysis. The findings suggest a need for further exploration of raster data structuring