ICFICE 2019
Current Result Document :
| 한글제목(Korean Title) |
An IoT-Based Object Detection and Alerting System for Livestock Disease Prevention |
| 영문제목(English Title) |
An IoT-Based Object Detection and Alerting System for Livestock Disease Prevention |
| 저자(Author) |
Wonseok Jung
Hyeon Park
Se-Han Kim
Jeongwook Seo
|
| 원문수록처(Citation) |
VOL 11 NO. 01 PP. 0337 ~ 0340 (2019. 06) |
한글내용
(Korean Abstract) |
|
영문내용
(English Abstract) |
In this paper, we implement object detection system for animal disease prevention through Faster Region-based Convolutional Neural Network (R-CNN) model and You Only Look Once (YOLO) v3 model. For object detection systems, we derive visual targets (pigs, human, trucks) and create open dataset through image collection and labeling. The open dataset is used to design the Faster R-CNN model for livestock disease detection of the object detection engine and the YOLOv3 model for farm environment detection of the object detection engine is designed using the pre-learned parameters. For the experiment, we use a webcam to capture the image of the visual target and detect the object using the designed model. The detected result is encoded for sharing. Then, the detection result is transmitted to the Internet of Things (IoT) server through the IoT client conforming to oneM2M standard. As a result, the Faster R-CNN model for animal disease detection was about 43.58%, and the YOLOv3 model for farm environment detection was about 55.17% and about 62.16%, respectively. The encoded data collected in the IoT server is decoded and sent to the registered users through the social network service (SNS) agent to implement the object detection system for the prevention of livestock diseases.
|
| 키워드(Keyword) |
Faster R-CNN
YOLOv3
Object Detection
Livestock Disease Prevention
Internet of Things
|
| 파일첨부 |
PDF 다운로드
|
