Improving precision analysis of power line modeling and integrating data processing chains for automating 3D building rooftop modeling
The project consists of two independent segments. The first part, to be managed by PhD student Yoonseok Jwa, aims to develop and evaluate new photogrammetric computer vision algorithms for detecting and identifying POAs and insulator types attached to power lines (PL). The expected benefit to the partner can be summarized as: (a) Efficiency- automatic detection and identification of POAs and insulator types help redirect resources to other operations and reduce the time needed for modeling, (b) Accuracy- more accurate clearance quantity which threaten PL systems, (c) Productivity- enhancing the updating process of a geospatial database in challenging environments and detecting potential encroachment dangers to the power lines. The second part, to be managed by PhD student Heungsik Kim, aims to design and develop a new automatic workflow for 3D building rooftop modeling by integrating state-of-the-art algorithms with current data processing chain. The expected benefit to the partner can be summarized as: (a) Efficiency- significant reduction in both time and cost required for reconstructing the 3D building models from the raw LiDAR data, (b) Accuracy-the proposed technique ensures that accurate 3D building models can be constructed automatically and free from human-induced errors, (c) Productivity- the new technique is expected to reduce the processing time to construct a 3D building model thereby allowing resources to be more focused on quality assurance and other operations.