Print Email Facebook Twitter The Smart Point Cloud framework to detect pipelines using raw point cloud generated from panoramic images Title The Smart Point Cloud framework to detect pipelines using raw point cloud generated from panoramic images Author Kotoula, Lydia (TU Delft Architecture and the Built Environment) Contributor Verbree, E. (mentor) van Oosterom, P.J.M. (graduation committee) Degree granting institution Delft University of Technology Programme Geomatics Date 2018-07-12 Abstract Spatial data acquisition is rapidly developing, making point clouds available and easily accessible for many applications and to many end-users. Nowadays, point clouds are the main surveying product and have more added value than derived products as they keep details and they are not interpolated. Raw point clouds do not contain information that relates the points to the semantic meanings of the real word objects that are represented. Moreover, through different procedures (classification and segmentation techniques) important semantic information could be derived, creating an intelligent environment and structure, a Smart Point Cloud (SPC). In this research, a SPC framework will be created combining different techniques and methods in order to detect the pipes in an industrial environment. Close-range photogrammetry will be used to generate a point cloud, for which panoramic images are the main source data. The features and the attributes from both the data (panoramic images and point cloud) will be combined to get characteristics from both sources (2D & 3D) in order to select, analyze, manipulate and identify the pipes as one object. Subject point cloudpanoramasobject identificationpipelines To reference this document use: http://resolver.tudelft.nl/uuid:40dddb4e-d0fa-42f5-9a54-f4fc32384cf4 Part of collection Student theses Document type master thesis Rights © 2018 Lydia Kotoula Files PDF 4625072_Thesis.pdf 16.52 MB PDF 4625072_P5_Presentation.pdf 3.84 MB Close viewer /islandora/object/uuid:40dddb4e-d0fa-42f5-9a54-f4fc32384cf4/datastream/OBJ1/view