Abstract

Computing the epipolar geometry between cameras withvery different viewpoints is often problematic as matchingpoints are hard to find. In these cases, it has been proposedto use information from dynamic objects in the scene forsuggesting point and line correspondences. We propose a speed up of about two orders of magnitude,as well as an increase in robustness and accuracy, to meth-ods computing epipolar geometry from dynamic silhouettes.This improvement is based on a new temporal signature:motion barcode for lines. Motion barcode is a binary tem-poral sequence for lines, indicating for each frame the ex-istence of at least one foreground pixel on that line. Themotion barcodes of two corresponding epipolar lines arevery similar, so the search for corresponding epipolar lines can be limited only to lines having similar barcodes. The use of motion barcodes leads to increased speed, accuracy, and robustness in computing the epipolar geometry.