Seg2Pose: Pose Estimations from Instance Segmentation Masks in One or Multiple Views for Traffic Applications

Martin Ahrnbom, Ivar Persson, Mikael Nilsson

2022

Abstract

A system we denote Seg2Pose is presented which converts pixel coordinate tracks, represented by instance segmentation masks across multiple video frames, into world coordinate pose tracks, for road users seen by static surveillance cameras. The road users are bound to a ground surface represented by a number of 3D points and does not necessarily have to be perfectly flat. The system works with one or more views, by using a late fusion scheme. An approximate position, denoted the normal position, is computed from the camera calibration, per-class default heights and the ground surface model. The position is then refined a novel Convolutional Neural Network we denote Seg2PoseNet, taking instance segmentations and cropping positioning as its input. We evaluate this system quantitatively both on synthetic data from CARLA Simulator and on a real recording from a trinocular camera. The system outperforms the baseline method of only using the normal positions, which is roughly equivalent of a typical 2D to 3D conversion system, in both datasets.

Paper Citation

in Harvard Style

Ahrnbom M., Persson I. and Nilsson M. (2022). Seg2Pose: Pose Estimations from Instance Segmentation Masks in One or Multiple Views for Traffic Applications. In Proceedings of the 17th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2022) - Volume 5: VISAPP; ISBN 978-989-758-555-5, SciTePress, pages 777-784. DOI: 10.5220/0010777700003124

in Bibtex Style

@conference{visapp22,
author={Martin Ahrnbom and Ivar Persson and Mikael Nilsson},
title={Seg2Pose: Pose Estimations from Instance Segmentation Masks in One or Multiple Views for Traffic Applications},
booktitle={Proceedings of the 17th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2022) - Volume 5: VISAPP},
year={2022},
pages={777-784},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0010777700003124},
isbn={978-989-758-555-5},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 17th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2022) - Volume 5: VISAPP
TI - Seg2Pose: Pose Estimations from Instance Segmentation Masks in One or Multiple Views for Traffic Applications
SN - 978-989-758-555-5
AU - Ahrnbom M.
AU - Persson I.
AU - Nilsson M.
PY - 2022
SP - 777
EP - 784
DO - 10.5220/0010777700003124
PB - SciTePress