3D Digital modelling

Teaching in lecture room, field data collection and labor experiments.

Knowledge:

The candidate should have knowledge of:

• the working principle of a laser scanner, the classifications and parameter settings, the error sources and calibrations
• the differences of laser scanning and photogrammetry
• the data format and characteristics of Laser scanning point clouds
• the process of data preprocessing including noise filtering, strip adjustment and data registration
• principles and algorithms of classifications and segmentations, and the object detection as well

Skills:

The candidate is able to:

• process the raw data acquired directly from an airborne laser scanner and terrestrial laser scanner
• model fittings by using RANSAC or improved RANSAC
• merge point clouds acquired from UAV borne and terrestrial laser scanner by using different registration methods
• adapt and apply filters from digital image process for 3D point clouds
• detect buildings, trees, street lights and power lines from 3D point clouds
• modelling the detected objects in CityGML

General competence:

The candidate can:

• understand the relations and transformations among different coordinate system induced by the sensors (GNSS, INS, Laser transmitter, scanner)
• classify 3D point clouds into several classes (bare earth, cars, buildings and other man-made objects, power lines, pole-like objects)
• segment the 3D point clouds of a building into individual plane or curved facets and model them in 3D either by using data-driven method or model-driven method
• generate CityGML data files for the 3D objects
• understand and use professional terminology within the discipline
• work independently and in team and take the necessary initiatives
• identify common fields between this discipline and other professional disciplines and be open for inter disciplinary approach and cooperation

Teaching in lecture room, field data collection and labor experiments.