Abstract | ||
---|---|---|
We consider the problem of foreground and background extraction from compressed-sensed (CS) surveillance video. We propose, for the first time in the literature, a principal component analysis (PCA) approach that computes the low-rank subspace of the background scene directly in the CS domain. Rather than computing the conventional L-2-norm-based principal components, which are simply the dominant left singular vectors of the CS measurement matrix, we compute the principal components under an L-1-norm maximization criterion. The background scene is then obtained by projecting the CS measurement vector onto the L-1 principal components followed by total-variation (TV) minimization image recovery. The proposed L-1-norm procedure directly carries out low-rank background representation without reconstructing the video sequence and, at the same time, exhibits significant robustness against outliers in CS measurements compared to L-2-norm PCA. |
Year | DOI | Venue |
---|---|---|
2016 | 10.1117/12.2179722 | COMPRESSIVE SENSING IV |
Keywords | Field | DocType |
Compressed sensing, convex optimization, feature extraction, L-1 principle component analysis, singular value decomposition, surveillance video, total-variation minimization | Computer vision,Subspace topology,Pattern recognition,Matrix (mathematics),Outlier,Robustness (computer science),Minification,Artificial intelligence,Image restoration,Maximization,Principal component analysis,Physics | Journal |
Volume | Issue | ISSN |
9484 | 3 | 0277-786X |
Citations | PageRank | References |
7 | 0.42 | 25 |
Authors | ||
2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Ying Liu | 1 | 60 | 5.05 |
Dimitris Pados | 2 | 208 | 26.49 |