3D Mosaicing Using Multiview Videos

Please enable javascript in your browser.

Taghaddosi, Mohsen | 2012

459 Viewed

Type of Document: M.Sc. Thesis
Language: Farsi
Document No: 43306 (19)
University: Sharif University of Technology
Department: Computer Engineering
Advisor(s): Kasaei, Shohreh
Abstract:
In many applications of multi view 3D reconstruction, image mosaicing is one of the most important processes. Image mosaicing is the action of stitching some pictures of a scene so that a complete picture of the scene is constructed. The proposed method is such that for PTZ cameras by extracting features from frame, the homography matrix between frames will be computed. Then by converting plane coordinate into spherical coordinate, a complete picture of the scene will be constructed and forground will be removed. For fixed cameras sufficient corresponding points will be selected for computing the homography matrix between cameras. Using the homography, the frames without forground will be rendered in arbitrary views. The main idea of this thesis is based on not decomposing homography matrix into its constructive matrices. Comparing this method to other methods in which decomposing homography matrix causes problems in mosaicing, our method is more robust. For removing forground and replacing their pixels in contrast to usual methods that use color features for forground detection, spatio-temporal features used in this thesis. Main advantage of this method is that it can be generalized and applied to other sports.
Keywords:
Image Mosaicking ; Background Modeling ; Soccer Math ; Multiview Geometry

No TOC