Surface Reconstruction of Point Clouds Captured with Microsoft Kinect

• Main Author: Hyvärinen, Juha
• Other Authors: Oulun seudun ammattikorkeakoulu
• Format: Bachelor Thesis
• Language: English
• Published: Oulun seudun ammattikorkeakoulu 2012
• Subjects: kolmiulotteisuus; virtuaalitodellisuus; Microsoft Kinect; realXtend; point cloud; Point Cloud Library; surface reconstruction; fi=Tieto- ja viestintätekniikka|sv=Informations- och kommunikationsteknik|en=Information and Communications Technology|; Tietotekniikan koulutusohjelma; Tundra
• Online Access: http://www.theseus.fi/handle/10024/42161

The main motivation behind this thesis was to create a new method of content creation for virtual spaces. This thesis focuses on how an already captured point cloud has to be processed for the optimal results of the object reconstruction. In this thesis realXtend Tundra, which is an open source framework for virtual spaces, is used as a target platform. The point cloud processing is integrated with Tundra in such a way that a user can import real world objects to the Tundra virtual space with an easy-to-use procedure. This thesis is based on the work done in the bachelor thesis of Miika Santala, who researched the capturing of real world objects to a point cloud by using Microsoft Kinect and the Point Cloud Library libraries (38). In this thesis, the open source Point Cloud Library was used for the point cloud processing. The thesis was commissioned by the Chiru project, Intel and Nokia Joint Innovation Center, Center for Internet Excellence, University of Oulu, which concentrates on researching and developing new 3D based user interaction paradigms. As a result, a working process for object capturing was achieved. A real world object can be captured to a point cloud and processed to the Ogre mesh, which can be added to a virtual space scene. The whole toolset is implemented to the realXtend Tundra framework as a module.