Geometrical and statistical incremental semantic modeling on mobile devices

Rafael Roberto, João Paulo Lima, Hideaki Uchiyama, Veronica Teichrieb, Rin ichiro Taniguchi

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Improvements on mobile devices allowed tracking applications to be executed on such platforms. However, there still remain several challenges in the field of mobile tracking, such as the extraction of high-level semantic information from point clouds. This task is more challenging when using monocular visual SLAM systems that output noisy sparse data. In this paper, we propose a primitive modeling method using both geometric and statistical analyses for sparse point clouds that can be executed on mobile devices. The main idea is to use the incremental mapping process of SLAM systems for analyzing the geometric relationship between the point cloud and the estimated shapes over time and selecting only reliably-modeled shapes. Besides that, a statistical evaluation that assesses if the modeling is random is incorporated to filter wrongly-detected primitives in unstable estimations. Our evaluation indicates that the proposed method was able to improve both precision and consistency of correct detection when compared with existing techniques. The mobile version execution is 8.5 to 9.9 times slower in comparison with the desktop implementation. However, it uses up to 30.5% of CPU load, which allows it to run on a separate thread, in parallel with the visual SLAM technique. Additional evaluations show that CPU load, energy consumption and RAM memory usage were not a concern when running our method on mobile devices.

Original languageEnglish
Pages (from-to)199-211
Number of pages13
JournalComputers and Graphics (Pergamon)
Volume84
DOIs
Publication statusPublished - Nov 2019

Fingerprint

Mobile devices
Semantics
Program processors
Random access storage
Energy utilization
Data storage equipment

All Science Journal Classification (ASJC) codes

  • Engineering(all)
  • Human-Computer Interaction
  • Computer Graphics and Computer-Aided Design

Cite this

Geometrical and statistical incremental semantic modeling on mobile devices. / Roberto, Rafael; Lima, João Paulo; Uchiyama, Hideaki; Teichrieb, Veronica; Taniguchi, Rin ichiro.

In: Computers and Graphics (Pergamon), Vol. 84, 11.2019, p. 199-211.

Research output: Contribution to journalArticle

@article{26bf0686d5174dbf833c1282ea58defb,
title = "Geometrical and statistical incremental semantic modeling on mobile devices",
abstract = "Improvements on mobile devices allowed tracking applications to be executed on such platforms. However, there still remain several challenges in the field of mobile tracking, such as the extraction of high-level semantic information from point clouds. This task is more challenging when using monocular visual SLAM systems that output noisy sparse data. In this paper, we propose a primitive modeling method using both geometric and statistical analyses for sparse point clouds that can be executed on mobile devices. The main idea is to use the incremental mapping process of SLAM systems for analyzing the geometric relationship between the point cloud and the estimated shapes over time and selecting only reliably-modeled shapes. Besides that, a statistical evaluation that assesses if the modeling is random is incorporated to filter wrongly-detected primitives in unstable estimations. Our evaluation indicates that the proposed method was able to improve both precision and consistency of correct detection when compared with existing techniques. The mobile version execution is 8.5 to 9.9 times slower in comparison with the desktop implementation. However, it uses up to 30.5{\%} of CPU load, which allows it to run on a separate thread, in parallel with the visual SLAM technique. Additional evaluations show that CPU load, energy consumption and RAM memory usage were not a concern when running our method on mobile devices.",
author = "Rafael Roberto and Lima, {Jo{\~a}o Paulo} and Hideaki Uchiyama and Veronica Teichrieb and Taniguchi, {Rin ichiro}",
year = "2019",
month = "11",
doi = "10.1016/j.cag.2019.09.003",
language = "English",
volume = "84",
pages = "199--211",
journal = "Computers and Graphics",
issn = "0097-8493",
publisher = "Elsevier Limited",

}

TY - JOUR

T1 - Geometrical and statistical incremental semantic modeling on mobile devices

AU - Roberto, Rafael

AU - Lima, João Paulo

AU - Uchiyama, Hideaki

AU - Teichrieb, Veronica

AU - Taniguchi, Rin ichiro

PY - 2019/11

Y1 - 2019/11

N2 - Improvements on mobile devices allowed tracking applications to be executed on such platforms. However, there still remain several challenges in the field of mobile tracking, such as the extraction of high-level semantic information from point clouds. This task is more challenging when using monocular visual SLAM systems that output noisy sparse data. In this paper, we propose a primitive modeling method using both geometric and statistical analyses for sparse point clouds that can be executed on mobile devices. The main idea is to use the incremental mapping process of SLAM systems for analyzing the geometric relationship between the point cloud and the estimated shapes over time and selecting only reliably-modeled shapes. Besides that, a statistical evaluation that assesses if the modeling is random is incorporated to filter wrongly-detected primitives in unstable estimations. Our evaluation indicates that the proposed method was able to improve both precision and consistency of correct detection when compared with existing techniques. The mobile version execution is 8.5 to 9.9 times slower in comparison with the desktop implementation. However, it uses up to 30.5% of CPU load, which allows it to run on a separate thread, in parallel with the visual SLAM technique. Additional evaluations show that CPU load, energy consumption and RAM memory usage were not a concern when running our method on mobile devices.

AB - Improvements on mobile devices allowed tracking applications to be executed on such platforms. However, there still remain several challenges in the field of mobile tracking, such as the extraction of high-level semantic information from point clouds. This task is more challenging when using monocular visual SLAM systems that output noisy sparse data. In this paper, we propose a primitive modeling method using both geometric and statistical analyses for sparse point clouds that can be executed on mobile devices. The main idea is to use the incremental mapping process of SLAM systems for analyzing the geometric relationship between the point cloud and the estimated shapes over time and selecting only reliably-modeled shapes. Besides that, a statistical evaluation that assesses if the modeling is random is incorporated to filter wrongly-detected primitives in unstable estimations. Our evaluation indicates that the proposed method was able to improve both precision and consistency of correct detection when compared with existing techniques. The mobile version execution is 8.5 to 9.9 times slower in comparison with the desktop implementation. However, it uses up to 30.5% of CPU load, which allows it to run on a separate thread, in parallel with the visual SLAM technique. Additional evaluations show that CPU load, energy consumption and RAM memory usage were not a concern when running our method on mobile devices.

UR - http://www.scopus.com/inward/record.url?scp=85072772029&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85072772029&partnerID=8YFLogxK

U2 - 10.1016/j.cag.2019.09.003

DO - 10.1016/j.cag.2019.09.003

M3 - Article

AN - SCOPUS:85072772029

VL - 84

SP - 199

EP - 211

JO - Computers and Graphics

JF - Computers and Graphics

SN - 0097-8493

ER -