Sparse codes as Alpha Matte

In this paper, image matting is cast as a sparse coding problem wherein the sparse codes directly give the estimate of the alpha matte. Hence, there is no need to use the matting equation that restricts the estimate of α from a single pair of foreground (F) and background (B) samples. A probabilistic segmentation provides a confidence value on the pixel belonging to F or B, based on which a dictionary is formed for use in sparse coding. This allows the estimate of α from more than just one pair of (F,B) samples. Experimental results on a benchmark dataset show the proposed method performs close to state-of-the-art methods.