Mosaic-based figure-ground segmentation along with static segmentation by mean shift

Authors:
Joel Sole;Yu Huang;Joan Llach
Affiliations:
Thomson Corporate Research, Princeton, NJ;Thomson Corporate Research, Princeton, NJ;Thomson Corporate Research, Princeton, NJ
Venue:
SIP '07 Proceedings of the Ninth IASTED International Conference on Signal and Image Processing
Year:
2007

Citing 6
Cited 0

Computing occluding and transparent motions

International Journal of Computer Vision
Matrix computations (3rd ed.)

Matrix computations (3rd ed.)
Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography

Communications of the ACM
Mean Shift: A Robust Approach Toward Feature Space Analysis

IEEE Transactions on Pattern Analysis and Machine Intelligence
Recognising Panoramas

ICCV '03 Proceedings of the Ninth IEEE International Conference on Computer Vision - Volume 2
Distinctive Image Features from Scale-Invariant Keypoints

International Journal of Computer Vision

Quantified Score

Hi-index	0.00

Visualization

Abstract

This work deals with the problem of background-foreground segmentation in video scenes. We propose an approach that makes use of feature extraction and matching, robust estimation, background mosaic generation, mosaic back-projection, and segmentation. We make contributions in each of these steps. SIFT features are extracted from the video and then, features between consecutive frames are matched. There are mismatches as well as errors in the feature location. To surmount this, we use a robust approach that employs a modified version of the RANSAC algorithm and weighted total least squares. Knowing the global motion allows creating an initial foreground segmentation and the generation of a mosaic mainly using background data. The moving object extraction is done by combining a mean shift-based frame segmentation along with the information given by the error of the mosaic back-projection into each frame.