Motion-aware temporal coherence for video resizing

Figure 1: Overview of our automatic content-aware video resizing framework. We align the original frames of a video clip to a common coordinate system by estimating interframe camera motion, so that corresponding components have roughly the same spatial coordinates. We achieve spatially and temporally coherent resizing of the aligned frames by preserving the relative positions of corresponding components within a grid-based optimization framework. The final resized video is reconstructed by transforming every video frame back to the original coordinate system. Temporal coherence is crucial in content-aware video retargeting. To date, this problem has been addressed by constraining tempo-rally adjacent pixels to be transformed coherently. However, due to the motion-oblivious nature of this simple constraint, the retargeted videos often exhibit flickering or waving artifacts, especially when significant camera or object motions are involved. Since the feature correspondence across frames varies spatially with both camera and object motion, motion-aware treatment of features is required for video resizing. This motivated us to align consecutive frames by estimating interframe camera motion and to constrain relative po-sitions in the aligned frames. To preserve object motion, we detect distinct moving areas of objects across multiple frames and con-strain each of them to be resized consistently. We build a com-plete video resizing framework by incorporating our motion-aware constraints with an adaptation of the scale-and-stretch optimization recently proposed by Wang and colleagues. Our streaming imple-mentation of the framework allows efficient resizing of long video sequences with low memory cost. Experiments demonstrate that our method produces spatiotemporally coherent retargeting results even for challenging examples with complex camera and object mo-tion, which are difficult to handle with previous techniques