Display options
Cite
Chen Y, Chan AB. Enhanced figure-ground classification with background prior propagation. IEEE Trans Image Process. 2015;24(3):873-85doi: 10.1109/TIP.2015.2389612.
Chen, Y., & Chan, A. B. (2015). Enhanced figure-ground classification with background prior propagation. IEEE transactions on image processing : a publication of the IEEE Signal Processing Society, 24(3), 873-85. https://doi.org/10.1109/TIP.2015.2389612
Chen, Yisong, and Chan, Antoni B. "Enhanced figure-ground classification with background prior propagation." IEEE transactions on image processing : a publication of the IEEE Signal Processing Society vol. 24,3 (2015): 873-85. doi: https://doi.org/10.1109/TIP.2015.2389612
Chen Y, Chan AB. Enhanced figure-ground classification with background prior propagation. IEEE Trans Image Process. 2015 Mar;24(3):873-85. doi: 10.1109/TIP.2015.2389612. Epub 2015 Jan 08. PMID: 25585417.
Copy Download .nbib Format: NLM
Share it on

IEEE Trans Image Process. 2015 Mar;24(3):873-85. doi: 10.1109/TIP.2015.2389612. Epub 2015 Jan 08.

Enhanced figure-ground classification with background prior propagation.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society

Yisong Chen, Antoni B Chan

PMID: 25585417 DOI: 10.1109/TIP.2015.2389612

Abstract

We present an adaptive figure-ground segmentation algorithm that is capable of extracting foreground objects in a generic environment. Starting from an interactively assigned background mask, an initial background prior is defined and multiple soft-label partitions are generated from different foreground priors by progressive patch merging. These partitions are fused to produce a foreground probability map. The probability map is then binarized via threshold sweeping to create multiple hard-label candidates. A set of segmentation hypotheses is formed using different evaluation scores. From this set, the hypothesis with maximal local stability is propagated as the new background prior, and the segmentation process is repeated until convergence. Similarity voting is used to select a winner set, and the corresponding hypotheses are fused to yield the final segmentation result. Experiments indicate that our method performs at or above the current state-of-the-art on several data sets, with particular success on challenging scenes that contain irregular or multiple-connected foregrounds.

Publication Types