Persons
Ing. et Ing. Václav Vávra
All publications
DoG Accuracy Via Equivariance: Get The Interpolation Right
- Authors: Ing. et Ing. Václav Vávra, Mgr. Dmytro Mishkin, Ph.D., prof. Ing. Jiří Matas, Ph.D.,
- Publication: 2023 IEEE International Conference on Image Processing (ICIP). New York: Institute of Electrical and Electronics Engineers, 2023. p. 136-140. ISBN 978-1-7281-9835-4.
- Year: 2023
- DOI: 10.1109/ICIP49359.2023.10222153
- Link: https://doi.org/10.1109/ICIP49359.2023.10222153
- Department: Visual Recognition Group
-
Annotation:
We study the influence of image interpolation algorithms on local feature detectors operating on a scale pyramid, focusing on the Difference-of-Gaussian, as used in SIFT. We show that commonly used implementations, such as in OpenCV and Kornia, are neither rotational nor scale equivariant. We present a simple solution and demonstrate its positive influence on the downstream image matching tasks. The implementation of the method has been accepted in standard libraries OpenCV and Kornia.
