The number of large infrastructure projects is increasing worldwide. When analyzing the often complex construction plans, construction planners and engineers face the difficulty of quickly and correctly capturing all the objects and materials depicted in them, which can lead to high follow-up costs and delays. By using artificial intelligence (AI) technologies in the field of computer vision, construction plans are to be analyzed faster and more reliably in the future.

As part of the research project, an approach is to be developed that can simultaneously segment individual components in the plan and detect and interpret the associated text labels. The core task of the IPF is the development and implementation of a corresponding robust automatic procedure based on computer vision and AI methods. The research work is carried out in cooperation with Röwaplan AG (www.roewaplan.de) and has a high practical relevance.