This paper introduces a novel approach to indoor visual localization based on a particle filter, CNN-segmentation and the nearest edge method for particle weight estimation. A main algorithm is used by detecting the edges on the image from camera and then mapping them to a3D model of the room. The main contribution of the paper is to introduce a novel approach that allows to get rid of such problems as noise generated by textured objects, edges created by dynamic objects and groups of unexpected objects which are considered to be excessive and can be removed from localization algorithm. The algorithm described in this paper was verified on a computer model in ROS/Gazebo environment.
Visual Localization for Copter based on 3D Model of Environment with CNN Segmentation
Full text:
pdf
Pre-published on:
February 26, 2018
Published on:
March 08, 2018
Abstract
DOI: https://doi.org/10.22323/1.300.0035
How to cite
Metadata are provided both in "article" format (very similar to INSPIRE) as this helps creating very compact bibliographies which can be beneficial to authors and readers, and in "proceeding" format which is more detailed and complete.
Open Access
Copyright owned by the author(s) under the term of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
Copyright owned by the author(s) under the term of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.