Title :
Correcting polyhedral projections for scene reconstruction
Author :
Ros, Lluis ; Thomas, Federico
Author_Institution :
Inst. de Robotica i Inf. Ind., CSIC, Barcelona, Spain
Abstract :
This paper presents a new algorithm for correcting incorrect projections of a polyhedral scene. Such projections arise in applications where an image of a polyhedral world is taken and its edges and vertices are extracted. Along the way, the true positions of the vertices in the 2D projection are perturbed due to digitization errors and the preprocessing. As most available algorithms for reconstructing polyhedral scenes from projections are "superstrict", they judge these noisy inputs as incorrect and fail to obtain a 3D scene from them. The method presented overcomes this problem by moving the positions of all vertices until a very close correct projection is found. With this tool, any superstrict method for reconstructing scenes from projections is now practical, as it can be applied to the corrected projection.
Keywords :
computational geometry; computer vision; image reconstruction; stereo image processing; 2D projection; 3D scene; image reconstruction; polyhedral projections; projection correction; scene reconstruction; superstrict method; Application software; Artificial intelligence; Computer vision; Humans; Image edge detection; Image reconstruction; Intelligent robots; Layout; Robot vision systems; Testing;
Conference_Titel :
Robotics and Automation, 2001. Proceedings 2001 ICRA. IEEE International Conference on
Print_ISBN :
0-7803-6576-3
DOI :
10.1109/ROBOT.2001.932921