V. Fremont and R. Chellali (France)
3D Reconstruction, 3D Visualization, Computer Graphics, Computer Vision, Sequence of Images, Conics Properties.
This paper describes a method that can produce 3D graphi cal models from real objects. Given an image sequence of a rotating complex shape object or viewed by a moving video camera under a fixed axis rotational motion, the presented algorithm generates automatically the 3D model. Rela tive to previous approaches, the algorithm described here is only based on conics properties and uses the spatiotempo ral aspect of the sequence of images. From the projective properties of the conics and using the camera calibration parameters, the metric 3D coordinates of a moving point are obtained from the geometric locus of the image points trajectories. An algorithm has been implemented to com pute the 3D reconstruction automatically. Both computer simulations and real image sequences have been used to test the proposed technique.
Important Links:
Go Back
