Scene-consistent detection of feature points in video sequences

Research output: Contribution to journalArticlepeer-review

6 Scopus citations

Abstract

Detection of feature points in images is an important preprocessing stage for many algorithms in Computer Vision. We address the problem of detection of feature points in video sequences of 3D scenes, which could be mainly used for obtaining scene correspondence. The main feature we use is the zero crossing of the intensity gradient argument. We analytically show that this local feature corresponds to specific constraints on the local 3D geometry of the scene, thus ensuring that the detected points are based on real 3D features. We present a robust algorithm that tracks the detected points along a video sequence, and suggest some criteria for quantitative evaluation of such algorithms. These criteria serve in a comparison of the suggested operator with four other feature trackers. The suggested criteria are generic and could serve other researchers as well for performance evaluation of stable point detectors.

Original languageEnglish
Pages (from-to)1-29
Number of pages29
JournalComputer Vision and Image Understanding
Volume97
Issue number1
DOIs
StatePublished - Jan 2005

Funding

FundersFunder number
Excellence Center for Geometric Computing of the Israel Academy of Science
Minerva Minkowski center for geometry
Ministry of Science and Technology, Israel

    Keywords

    • Feature point detection
    • Scene-consistent detection
    • Stable point tracking
    • Tracking evaluation

    Fingerprint

    Dive into the research topics of 'Scene-consistent detection of feature points in video sequences'. Together they form a unique fingerprint.

    Cite this