Optimized Color Sampling for Robust Matting

By Jue Wang, Michael F. Cohen. 2007 IEEE Conference on Computer Vision and Pattern Recognition, 2007

459 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/dbf7c43877a418bf0fd09c8f9ef034410c92a30d

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

Geodesic Matting: A Framework for Fast Interactive Image and Video Segmentation and Matting (2009, background): “The authors of [29] proposed an improved color sampling method for natural image matting, and demonstrated very good performance.”
A perceptually motivated online benchmark for image matting (2009, methodology): “Then we computed error rates (mean squared error ) for all images with a variety of matting algorithms (i.e. [23] [14] [9] [20] [6] [8]).”
Semantic Object Selection (2014, methodology): “Matting methods [5, 14, 27] require a trimap to be specified.”
Image matting for fusion of multi-focus images in dynamic scenes (2013, methodology): “In order to obtain the focused region of each source image, taking the trimap Tn obtained above as one input, the robust image matting algorithm [24] is performed on source image In to get the alpha matte anðx; yÞ.”
A global sampling method for alpha matting (2011, methodology): “[12] collected samples nearby in geodesic distance, as opposed to the Euclidean distance used in [19].”
A Geodesic Framework for Fast Interactive Image and Video Segmentation and Matting (2007, background): “The authors of [19] proposed an improved color sampling method for natural image matting, and demonstrated very good performance.”
KNN Matting (2013, unknown): “Noting that relevant color sampling improves performance [27], [6], this approach samples and matches in a randomized manner relevant nonlocal neighbors in a joint foreground-background space which, as mentioned, can become prohibitively large if it is 2176 IEEE TRANSACTIONS ON PATTERN ANALYSIS AND ”