Panning and Jitter Invariant Incremental Principal Component Pursuit for Video Background Modeling

Figure 7

Selected frames from three sequences of the Davis dataset (Section 4.3) along with the segmented sparse components for incPCP-PTI, DECOLOR, and LTVA (aPP and mPP). The sequences in which DECOLOR did not work properly are left blank.