A novel method of object detection from a moving camera based on image matching and frame coupling

Yong Chen; Rong hua Zhang; Lei Shang

doi:10.1371/journal.pone.0109809

A novel method of object detection from a moving camera based on image matching and frame coupling

PLoS One. 2014 Oct 29;9(10):e109809. doi: 10.1371/journal.pone.0109809. eCollection 2014.

Authors

Yong Chen¹, Rong hua Zhang¹, Lei Shang¹

Affiliation

¹ School of Energy Science and Engineering, University of Electronic Science and Technology of China, Chengdu, China.

Abstract

A new method based on image matching and frame coupling to handle the problems of object detection caused by a moving camera and object motion is presented in this paper. First, feature points are extracted from each frame. Then, motion parameters can be obtained. Sub-images are extracted from the corresponding frame via these motion parameters. Furthermore, a novel searching method for potential orientations improves efficiency and accuracy. Finally, a method based on frame coupling is adopted, which improves the accuracy of object detection. The results demonstrate the effectiveness and feasibility of our proposed method for a moving object with changing posture and with a moving camera.

Publication types

Research Support, Non-U.S. Gov't

MeSH terms

Algorithms*
Image Interpretation, Computer-Assisted* / methods
Motion

Grants and funding

Funding was provided by National Natural Science Foundation of China under grant (61105030), The Fundamental Research Funds for the Central Universities (ZYGX2011J021), and the Scientific and Technical Supporting Programs of Sichuan Province (2013GZ0054). The funders had a role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.