Multiple Frames Matching for Object Discovery in Video

Otilia Stretcu and Marius Leordeanu

Abstract

Automatic discovery of foreground objects in video sequences is an important problem in computer vision with applications to object tracking, video segmentation and classification. We propose an efficient method for the discovery of object bounding boxes and the corresponding soft-segmentation masks across multiple video frames. We offer a graph matching formulation for bounding box selection and refinement using second and higher order terms. Our objective function takes into consideration local, frame-based information, as well as spatiotemporal and appearance consistency over multiple frames. First, we find an initial pool of candidate boxes using a novel and fast foreground estimation method in video, based on Principal Component Analysis. Then, we match the boxes across multiple frames using pairwise geometric and appearance terms. Finally, we refine their location and soft-segmentation using higher order potentials that establish appearance regularity over multiple frames. We test our method on the large scale YouTube-Objects dataset and obtain state-of-the-art results on several object classes.

Session

Detection and Recognition

Files

Extended Abstract (PDF, 229K)
Paper (PDF, 3M)

DOI

10.5244/C.29.186
https://dx.doi.org/10.5244/C.29.186

Citation

Bibtex

@inproceedings{BMVC2015_186,
	title={Multiple Frames Matching for Object Discovery in Video},
	author={Otilia Stretcu and Marius Leordeanu},
	year={2015},
	month={September},
	pages={186.1-186.12},
	articleno={186},
	numpages={12},
	booktitle={Proceedings of the British Machine Vision Conference (BMVC)},
	publisher={BMVA Press},
	editor={Xianghua Xie, Mark W. Jones, and Gary K. L. Tam},
	doi={10.5244/C.29.186},
	isbn={1-901725-53-7},
	url={https://dx.doi.org/10.5244/C.29.186}
}