Version 1
: Received: 24 March 2020 / Approved: 25 March 2020 / Online: 25 March 2020 (08:57:29 CET)
Version 2
: Received: 9 September 2020 / Approved: 13 September 2020 / Online: 13 September 2020 (11:16:17 CEST)
How to cite:
Kaihoko, Y.; Tan, P.X.; Kamioka, E. Prevention of Unintended Appearance in Photos Based on Human Behaviors Analysis. Preprints2020, 2020030376. https://doi.org/10.20944/preprints202003.0376.v1
Kaihoko, Y.; Tan, P.X.; Kamioka, E. Prevention of Unintended Appearance in Photos Based on Human Behaviors Analysis. Preprints 2020, 2020030376. https://doi.org/10.20944/preprints202003.0376.v1
Kaihoko, Y.; Tan, P.X.; Kamioka, E. Prevention of Unintended Appearance in Photos Based on Human Behaviors Analysis. Preprints2020, 2020030376. https://doi.org/10.20944/preprints202003.0376.v1
APA Style
Kaihoko, Y., Tan, P.X., & Kamioka, E. (2020). Prevention of Unintended Appearance in Photos Based on Human Behaviors Analysis. Preprints. https://doi.org/10.20944/preprints202003.0376.v1
Chicago/Turabian Style
Kaihoko, Y., Phan Xuan Tan and Eiji Kamioka. 2020 "Prevention of Unintended Appearance in Photos Based on Human Behaviors Analysis" Preprints. https://doi.org/10.20944/preprints202003.0376.v1
Abstract
Many people can take photos with smartphones and easily post photos via SNS (Social Network Services). This has caused a social problem that unintended appearance in photos may threaten the privacy of photographed persons. For this issue, numerous studies have already been introduced to prevent the unintended appearance in photos from the photographer’s side, but only a few methods tackled this from the photographed person's side. Therefore, we considered calling attention to a situation that a photo-taking behavior by a photographer can be automatically detected by using a wearable camera worn by a photographed person. In this paper, we propose an approach to detect photo-taking behaviors in video data taken from the wearable camera, analyzing specific human skeleton information. OpenPose is utilized to obtain the human’s skeleton information and the time-series data are analyzed. In addition, we compare two similar behaviors which are photo-taking behaviors and net-surfing behaviors. These video data are distinguished by DP matching and cross-validation. Finally, it is concluded that the detection accuracy of photo-taking behaviors is about 92.5%, which is satisfactory enough for this research purpose.
Computer Science and Mathematics, Information Systems
Copyright:
This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.