Removing Spinning Motion from Spherical Video Using a Rotationally Invariant Feature Extractor

Files in This Item:
Access to this item has been restricted by the copyright holder until:2019-10-18
File Description SizeFormat 
Final submit Removing_Spinning_Motion_from_Spherical_Video_using_a_Rotationally_Invariant_Feature_Extractor.pdf2.97 MBAdobe PDF    Request a copy
Title: Removing Spinning Motion from Spherical Video Using a Rotationally Invariant Feature Extractor
Authors: Slevin, MattCuffe, Paul
Permanent link: http://hdl.handle.net/10197/10985
Date: 17-Oct-2019
Online since: 2019-08-13T07:12:01Z
Abstract: This paper presents a method for removing rotational motion from an omnidirectional video captured from the viewpoint of a spinning object. For instance, the technique could be used to recover stable footage from spherical video captured from within a football used for competitive sports. The methodology uses a rotationally-invariant algorithm to obtain feature points and descriptors for pairs of successive frames. This information is then abstracted into three-dimensional point clouds, from which the Kabsch algorithm can infer the rotational motion between frames. The resulting rotational matrices are used to remap each equirectangular frame to a reference frame, and thereby offset the effect of frame-to-frame camera rotations. The algorithm has been successfully tested with various styles of footage. Additionally, the codebase implementing our algorithm has been freely shared through an open online repository.
Type of material: Conference Publication
Publisher: IEEE
Keywords: Omnidirectional VideoBall CameraVideo DerotationRotating Equirectangular Image
Language: en
Status of Item: Peer reviewed
Conference Details: IECON 2019, 45th Annual Conference of the IEEE Industrial Electronics Society (IES), Lisbon, Portugal, October 14-17 2019
Appears in Collections:Electrical and Electronic Engineering Research Collection

Show full item record

Google ScholarTM

Check


This item is available under the Attribution-NonCommercial-NoDerivs 3.0 Ireland. No item may be reproduced for commercial purposes. For other possible restrictions on use please refer to the publisher's URL where this is made available, or to notes contained in the item itself. Other terms may apply.