A System Framework for Localization and Mapping Using High Resolution Cameras of Mobile Devices
Abstract
We propose a hierarchical framework for processing high-resolution images on mobile devices for visual SLAM. It is based on the insights from analysis of new progress in primary features' detection, object detection and pose estimation. A rectification/unwarping operation is applied in regions of interest (ROIs) to improve the object/parts classification/detection performance; the object-part spatial relationships are created and contribute in map building, object detection, and localization; and a geometric constraints based pose refinement is followed to further improve the localization accuracy. Our design facilitates the more accurate pose estimating and localization using mobile devices for SLAM, and Augmented Reality/Mixed Reality applications.
Cite
Text
Liu and Li. "A System Framework for Localization and Mapping Using High Resolution Cameras of Mobile Devices." IEEE/CVF International Conference on Computer Vision Workshops, 2019. doi:10.1109/ICCVW.2019.00472Markdown
[Liu and Li. "A System Framework for Localization and Mapping Using High Resolution Cameras of Mobile Devices." IEEE/CVF International Conference on Computer Vision Workshops, 2019.](https://mlanthology.org/iccvw/2019/liu2019iccvw-system/) doi:10.1109/ICCVW.2019.00472BibTeX
@inproceedings{liu2019iccvw-system,
title = {{A System Framework for Localization and Mapping Using High Resolution Cameras of Mobile Devices}},
author = {Liu, Lifeng and Li, Jian},
booktitle = {IEEE/CVF International Conference on Computer Vision Workshops},
year = {2019},
pages = {3796-3803},
doi = {10.1109/ICCVW.2019.00472},
url = {https://mlanthology.org/iccvw/2019/liu2019iccvw-system/}
}