Technical Report CIS9710

Title: Auto Photometric/Geometric Stereo from a Single Image or Class Based Reconstruction of Bilaterally Symmetric Objects
Authors: Ilan Shimshoni Yael Moses Michael Lindenbaum
Abstract: This paper proposes a class based approach to the reconstruction of the shape of 3D bilaterally symmetric objects from a single image. The special geometric and photometric constraints available from images of symmetric objects are derived and the evidence they carry of the 3D object shape is analyzed. If the correspondence between symmetric points is available then the photometric information is used to infer the object. If this correspondence is not available, then 3D reconstruction requires a combined form of both types constraints. The important characteristics of the algorithm are that unlike other shape recovery algorithms from a single image, this algorithm utilizes a local constraint which has to be satisfied at each point in the image to guarantee correct reconstruction. This local constraint is also used to recover exactly the global viewing and illumination parameters pose, illumination direction, and albedo. A useful byproduct of the reconstruction process is the correspondence function which is a relation between the image position of symmetric points.
CopyrightThe above paper is copyright by the Technion, Author(s), or others. Please contact the author(s) for more information

Remark: Any link to this technical report should be to this page (, rather than to the URL of the PDF files directly. The latter URLs may change without notice.

To the list of the CIS technical reports of 1997
To the main CS technical reports page

Computer science department, Technion