Determination of sign of gaussian curvature of surface having general reflectance property

Takayuki Okatani, Koichiro Deguchi

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review


A new method for determining sign of the Gaussian curvature at points on the surface of an object using images taken under three different lighting directions is presented. A smooth surface is segmented into two regions based on the sign of Gaussian curvature. The boundaries of these two regions are independent of the orientation of the object in 3D space and the position of the viewer. Thus they are useful for various applications such as object recognition and pose estimation. The present method does not require knowledge of the directions and strengths of the light sources except the rotation orientation of the directions about the viewing direction if they are linearly independent. It is applicable for almost any kind of surface reflecting properties.

Original languageEnglish
Title of host publicationComputer Vision - ACCV 1998 - 3rd Asian Conference on Computer Vision, Proceedings
EditorsRoland Chin, Ting-Chuen Pong
PublisherSpringer Verlag
Number of pages8
ISBN (Print)3540639306, 9783540639305
Publication statusPublished - 1997
Event3rd Asian Conference on Computer Vision, ACCV 1998 - Hong Kong, China
Duration: 1998 Jan 81998 Jan 10

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Conference3rd Asian Conference on Computer Vision, ACCV 1998
CityHong Kong


Dive into the research topics of 'Determination of sign of gaussian curvature of surface having general reflectance property'. Together they form a unique fingerprint.

Cite this