Skip to main navigation Skip to search Skip to main content

3D Keypoint Detection of Lying Human Body Using an RGB-D Camera

  • Harbin Institute of Technology Shenzhen

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

3D keypoint detection of lying human body is important to improve the efficiency of mobile rescue robots at the casualty collection point after a disaster (e.g., earthquake and mudslide). In this paper, we propose an efficient method for 3D human keypoint detection in a lying posture using an RGB-D camera. First, we use the current 2D human pose estimation algorithm for RGB images to obtain 2D keypoints of the whole body. We then obtain the final 3D coordinates of the human keypoint by processing the 2D coordinates through a filter of our design and combining the depth information with a coordinate transformation. Experiments show that the proposed method is accurate and fast enough to be used for 3D keypoint detection of lying human body by the mobile rescue robot at the casualty collection point.

Original languageEnglish
Title of host publication2023 42nd Chinese Control Conference, CCC 2023
PublisherIEEE Computer Society
Pages7459-7464
Number of pages6
ISBN (Electronic)9789887581543
DOIs
StatePublished - 2023
Externally publishedYes
Event42nd Chinese Control Conference, CCC 2023 - Tianjin, China
Duration: 24 Jul 202326 Jul 2023

Publication series

NameChinese Control Conference, CCC
Volume2023-July
ISSN (Print)1934-1768
ISSN (Electronic)2161-2927

Conference

Conference42nd Chinese Control Conference, CCC 2023
Country/TerritoryChina
CityTianjin
Period24/07/2326/07/23

Keywords

  • Deep Learning
  • Human Keypoint Detection
  • Mobile Rescue Robot
  • RGB-D

Fingerprint

Dive into the research topics of '3D Keypoint Detection of Lying Human Body Using an RGB-D Camera'. Together they form a unique fingerprint.

Cite this