Edgel sampling for edge-based tracking

摘要

Embodiments include selecting edgels for edge based tracking by dividing a reference image frame (RF) into N×M bins of pixels and projecting a subset of the edgels per bin into a current image frame (CF) using an estimated pose to identify valid bins of the RF as bins having their projected one edgel found within the borders of the CF. Then, K edgels of RF with different orientations from each valid bins may be selected. Then, the selected RF edgels of bins may be reduced by removing bins randomly, or first removing bins from the center of the RF of the image (then next removing the next further outward bins), until a desirable edgel number is obtained. Edge-based tracking can then be performed using the desirable edgel number, to track edges in current frame that are found in prior frame.

主权项

1. A machine implemented method comprising:
obtaining a reference image comprising an object to be tracked; detecting, in the reference image, a plurality of edgels of the object to be tracked; dividing the reference image into a plurality of bins; determining one or more valid bins from the plurality of bins in the reference image, wherein, each valid bin contains a corresponding at least one edgel of the object to be tracked that corresponds to an edgel of a tracked object in a current image captured subsequent to the reference image; and selecting, from each valid bin in the reference image, a corresponding subset of the plurality of edgels of the object to be tracked, wherein, in each subset with a plurality of edgels to be tracked, edgels of the object to be tracked with differing angular orientations are favored for selection in the subset.