Research On The Project Light Model In The Acquisition Of Three-Dimensional Shape

The computer acquisition of three-dimensional shape is becoming increasingly important commercially, with applications in fields such as computer vision, virtual and augmented reality, robot navigation, and manufacturing. After an analysis of the underlying assumptions of extant structured light techniques, a new light project model is presented for range scanning of rigid objects more rapidly. The dissertation focuses on the modeling and implementation of the system .The main content is depicted as follows:Firstly, a general introduction is given on the research of the history and development of three dimensional scene geometries. The methods that obtain depth via triangulation are systematically classified from the viewpoint of scanning way and are compared.Secondly, a common model is proposed for structure-lighted system, which is based on the theories of signal proccessing and becomes the basement of the illustration for the spatial deformed grating pattern and the spatial-modulation of projected light pattern. Then a unifying framework is given for depth from triangulation. Based on this framework, the fundamental factor of the acquisition system is that determining the correct correspondence between viewpoints of the scene is the fundamental factor. The various light-coding methods based on the framework, such as spatial coding model, temporal coding model and space-time coding model, are distinguished. Then the dissertation derives a new set of illumination patterns based on coding the boundaries between projected stripes. Thirdly, according to the environment of scene for scanning and the aim of acquisition, I describe an implementation that integrates these new codes with some special algorithms for tracking stripe boundaries and determining depths. These algorithms are used in the flow of the systems, such as stripe boundary code designing, the structured light pattern projection, the video capturing in Windows, the segmentation, stripe matching and decoding.Finally, the results of the acquisition are got after testing the implementation on serial scenes. According to the results, the advantages of the project light model are shown, as well as the limitations of the system. And some ideas for stripe code designing and structured light acquisition are presented, that might be focused in the future work.