TY - JOUR
T1 - Scene-consistent detection of feature points in video sequences
AU - Tankus, Ariel
AU - Yeshurun, Yehezkel
N1 - Funding Information:
We thank Professor Ronny Kimmel for his helpful remarks on level sets. The original image of Fig. 2B is courtesy of www.freeimages.co.uk . The original image of Fig. 2C is courtesy of the US Department of Agriculture. This research was supported by grants from the Minerva Minkowski center for geometry, the Excellence Center for Geometric Computing of the Israel Academy of Science, and Israel Ministry of Science.
PY - 2005/1
Y1 - 2005/1
N2 - 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.
AB - 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.
KW - Feature point detection
KW - Scene-consistent detection
KW - Stable point tracking
KW - Tracking evaluation
UR - http://www.scopus.com/inward/record.url?scp=11144225133&partnerID=8YFLogxK
U2 - 10.1016/j.cviu.2004.05.003
DO - 10.1016/j.cviu.2004.05.003
M3 - ???researchoutput.researchoutputtypes.contributiontojournal.article???
AN - SCOPUS:11144225133
SN - 1077-3142
VL - 97
SP - 1
EP - 29
JO - Computer Vision and Image Understanding
JF - Computer Vision and Image Understanding
IS - 1
ER -