Structure-from-Motion using Dense CNN Features with Keypoint Relocalization

  • 2018-05-10 08:35:06
  • Aji Resindra W, Akihiko Torii, Masatoshi Okutomi
  • 29


Structure from Motion (SfM) using imagery that involves extreme appearancechanges is yet a challenging task due to a loss of feature repeatability. Usingfeature correspondences obtained by matching densely extracted convolutionalneural network (CNN) features significantly improves the SfM reconstructioncapability. However, the reconstruction accuracy is limited by the spatialresolution of the extracted CNN features which is not even pixel-level accuracyin the existing approach. Providing dense feature matches with precise keypointpositions is not trivial because of memory limitation and computational burdenof dense features. To achieve accurate SfM reconstruction with highlyrepeatable dense features, we propose an SfM pipeline that uses dense CNNfeatures with relocalization of keypoint position that can efficiently andaccurately provide pixel-level feature correspondences. Then, we demonstrate onthe Aachen Day-Night dataset that the proposed SfM using dense CNN featureswith the keypoint relocalization outperforms a state-of-the-art SfM (COLMAPusing RootSIFT) by a large margin.


Introduction (beta)



Conclusion (beta)