Are you an EPFL student looking for a semester project?
Work with us on data science and visualisation projects, and deploy your project as an app on top of GraphSearch.
We propose a method to compute scale invariant features in omnidirectional images. We present a formulation based on Riemannian geometry for the definition of differential operators on non-Euclidian manifolds that describe the mirror and lens structure in omnidirectional imaging. These operators lead to a scale-space analysis that preserves the geometry of the visual information in omnidirectional images. We then build a novel scale-invariant feature detection framework for any omnidirectional image that can be mapped on the sphere. We also present a new descriptor and feature matching solution for omnidirectional images. The descriptor builds on the log-polar planar descriptors and adapts the descriptor computation to the specific geometry and the non-uniform sampling density of spherical and omnidirectional images. We further propose a rotation-invariant matching method that eliminates the orientation computation during the feature detection phase and thus decreases the computational complexity. Finally, we show that the proposed framework also permits to match features in images with different geometries. Experimental results demonstrate that the new feature detection method combined with the proposed descriptors offers promising performance and improves on the common SIFT features computed on the planar omnidirectional images as well as other state-of-the art methods for omnidirectional images.
Olga Fink, Ismail Nejjar, Mengjie Zhao
Martin Vetterli, Ivan Dokmanic, Puoya Tabaghi
Katrin Beyer, Bryan German Pantoja Rosero, Savvas Saloustros