Optical Flow Estimation Using Wavelet Motion Model

A motion estimation algorithm using wavelet approximation as an optical ow model has been developed to estimate accurate dense optical ow from an image sequence. This wavelet motion model is particularly useful in estimating optical ows with large displacement. Traditional pyramid methods which use the coarse-tone image pyramid by image burring in estimating optical ow often produce incorrect results when the coarse-level estimates contain large errors that cannot be corrected at the subsequent ner levels. This happens when regions of low texture become at or certain patterns result in spatial aliasing due to image blurring. Our method, in contrast, uses large-to-small fullresolution regions without blurring images, and simultaneously optimizes the coarser and ner parts of optical ow so that the large and small motion can be estimated correctly. We compare results obtained by using our method with those obtained by using one of the leading optical ow methods, the Szeliski pyramid spline-based method. The experiments include cases of small displacement (less than 4 pixels under 128 128 image size or equivalent displacement under other image sizes), and those of large displacement (10 pixels). While both methods produce comparable results when the displacements are small, our method outperforms pyramid spline-based method when the displacements are large.