Faculty Information |
|
Language | English |
The announcement title | A Preliminary Study on Automatic Recognition of Bedside Monitor Output Enhanced by
YOLO with Fisheye Camera |
Academic Society name | IEEE EMBC 2023 |
Promoters | IEEE |
Conference Type | International Society |
Announcement form | Poster presentation |
Presentation Type | General Lecture |
Lecture Type | 一般 |
Publisher and common publisher | Rumi Iwai, Takunori Shimazaki, Yoshifumi Kawakubo, Jun Mitsudo, Yuhei Hayashi, Shingo Ata, Takeshi Yokoyama,Daisuke Anzai |
Date | 2023/07/26 |
Country | Australia |
Venue (city and name of the country) |
Sydney |
Holding period | 2023/07/24~2023/07/27 |
Description | This study investigated the effectiveness of automatic recognition on bedside monitor
output by combining a fisheye camera and you only look once (YOLO), which is one of the real-time object detection algorithms, as a preliminary study. First, with a fisheye camera placed above the monitor, the developed system adjusted the distortion of the acquired image file. Then, the YOLO algorithm was applied to recognize the monitor output, including waveforms and numbers on the axes. The evaluation result was obtained as mAP (mean average precision) = 0.84, which clearly demonstrated the effectiveness of the image quantification of the bedside monitor for supporting remote area medicine. |