%0 Conference Paper %1 10328368 %A Seibt, Simon %A Von Rymon Lipinski, Bartosz %A Chang, Thomas %A Latoschik, Marc Erich %B 2023 IEEE International Conference on Image Processing Challenges and Workshops (ICIPCW) %D 2023 %K myown %P 1-5 %R 10.1109/ICIPC59416.2023.10328368 %T DFM4SFM - Dense Feature Matching for Structure from Motion %U https://ieeexplore.ieee.org/document/10328368 %X Structure from motion (SfM) is a fundamental task in computer vision and allows recovering the 3D structure of a stationary scene from an image set. Finding robust and accurate feature matches plays a crucial role in the early stages of SfM. So in this work, we propose a novel method for computing image correspondences based on dense feature matching (DFM) using homographic decomposition: The underlying pipeline provides refinement of existing matches through iterative rematching, detection of occlusions and extrapolation of additional matches in critical image areas between image pairs. Our main contributions are improvements of DFM specifically for SfM, resulting in global refinement and global extrapolation of image correspondences between related views. Furthermore, we propose an iterative version of the Delaunay-triangulation-based outlier detection algorithm for robust processing of repeated image patterns. Through experiments, we demonstrate that the proposed method significantlv improves the reconstruction accuracy.

@inproceedings{10328368, abstract = {Structure from motion (SfM) is a fundamental task in computer vision and allows recovering the 3D structure of a stationary scene from an image set. Finding robust and accurate feature matches plays a crucial role in the early stages of SfM. So in this work, we propose a novel method for computing image correspondences based on dense feature matching (DFM) using homographic decomposition: The underlying pipeline provides refinement of existing matches through iterative rematching, detection of occlusions and extrapolation of additional matches in critical image areas between image pairs. Our main contributions are improvements of DFM specifically for SfM, resulting in global refinement and global extrapolation of image correspondences between related views. Furthermore, we propose an iterative version of the Delaunay-triangulation-based outlier detection algorithm for robust processing of repeated image patterns. Through experiments, we demonstrate that the proposed method significantlv improves the reconstruction accuracy.}, added-at = {2023-12-13T07:47:18.000+0100}, author = {Seibt, Simon and Von Rymon Lipinski, Bartosz and Chang, Thomas and Latoschik, Marc Erich}, biburl = {https://www.bibsonomy.org/bibtex/2d1a58ae63c0a32d9b8888b4552fd7fe4/hci-uwb}, booktitle = {2023 IEEE International Conference on Image Processing Challenges and Workshops (ICIPCW)}, doi = {10.1109/ICIPC59416.2023.10328368}, interhash = {0d3a7196b7477d466a20ce3d21d3fe51}, intrahash = {d1a58ae63c0a32d9b8888b4552fd7fe4}, keywords = {myown}, pages = {1-5}, timestamp = {2023-12-13T07:47:18.000+0100}, title = {DFM4SFM - Dense Feature Matching for Structure from Motion}, url = {https://ieeexplore.ieee.org/document/10328368}, year = 2023 }