What can be determined from a full and a weak perspective image?

This paper presents a first investigation on the structure from motion problem from the combination of full and weak perspective images. This problem arises in multiresolution object modeling where multiple zoomed-in or close-up views are combined with wider or distant reference views. The narrow field-of-view (FOV) images from the zoomed-in or closeup views can be approximated as weak perspective projection. Using a full perspective projection model for the narrow FOV images, although more accurate, actually leads to instabilities during the estimation process due to the non-linearities in the imaging model. The weak perspective approximation leads to more stable estimation algorithms, although at the cost of a small amount of modeling inaccuracy. Previous work in structure from motion focused either on two (or more) perspective images or on a set of weak perspective (more generally, affine) images. The main contribution of this paper is the study of the SFM problem for the much neglected case of one perspective and one (or more) weak perspective image. We show that in contrast to the case of a pair of weak perspective images, there is adequate information to recover Euclidean structure from a single perspective and a single weak perspective image. The epipolar geometry is simpler than with two perspective images leading to simpler and more stable estimation algorithms. Computer simulation shows that more stable results can be obtained with the technique presented in this paper than if two images are both considered to be full perspective.

[1]  Peter Spellucci,et al.  An SQP method for general nonlinear programs using only equality constrained subproblems , 1998, Math. Program..

[2]  Olivier D. Faugeras,et al.  What can be seen in three dimensions with an uncalibrated stereo rig , 1992, ECCV.

[3]  Thomas S. Huang,et al.  Motion and structure from feature correspondences: a review , 1994, Proc. IEEE.

[4]  S. Ullman.  The Interpretation of Visual Motion , 1979 .

[5]  Thomas S. Huang,et al.  Theory of Reconstruction from Image Motion , 1992 .

[6]  Rajiv Gupta,et al.  Stereo from uncalibrated cameras , 1992, Proceedings 1992 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[7]  Richard I. Hartley,et al.  Euclidean Reconstruction from Uncalibrated Views , 1993, Applications of Invariance in Computer Vision.

[8]  Olivier D. Faugeras,et al.  Some Properties of the E Matrix in Two-View Motion Estimation , 1989, IEEE Trans. Pattern Anal. Mach. Intell..

[9]  J. Roach,et al.  Determining the movement of objects from a sequence of images , 1980, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[10]  Gene H. Golub,et al.  Matrix computations , 1983 .

[11]  Richard I. Hartley,et al.  In defence of the 8-point algorithm , 1995, Proceedings of IEEE International Conference on Computer Vision.

[12]  Larry S. Shapiro,et al.  Affine Analysis of Image Sequences: Contents , 1995 .

[13]  Yiannis Aloimonos,et al.  Perspective approximations , 1990, Image Vis. Comput..

[14]  Jake K. Aggarwal,et al.  On the computation of motion from sequences of images-A review , 1988, Proc. IEEE.

[15]  Amnon Shashua,et al.  Projective Structure from Uncalibrated Images: Structure From Motion and Recognition , 1994, IEEE Trans. Pattern Anal. Mach. Intell..

[16]  Gang Xu,et al.  Epipolar Geometry in Stereo, Motion and Object Recognition , 1996, Computational Imaging and Vision.

[17]  O. Faugeras.  Three-dimensional computer vision: a geometric viewpoint , 1993 .

[18]  Long Quan,et al.  Relative 3D Reconstruction Using Multiple Uncalibrated Images , 1995, Int. J. Robotics Res..

[19]  M. Tistarelli,et al.  Analysis of Image Sequences , 1985 .

[20]  S. Nash,et al.  Linear and Nonlinear Programming , 1987 .

[21]  Long Quan,et al.  Relative 3D Reconstruction Using Multiple Uncalibrated Images , 1993, Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.