ETRI-Knowledge Sharing Plaform

KOREAN
논문 검색
Type SCI
Year ~ Keyword

Detail

Conference Paper Deep Learning based Moving Object Detection using Portable Camera
Cited - time in scopus Share share facebook twitter linkedin kakaostory
Authors
Jaemin Cho, Sangseung Kang, Kye Kyung Kim
Issue Date
2019-06
Citation
International Conference on Ubiquitous Robots (UR) 2019, pp.1-2
Publisher
IEEE
Language
English
Type
Conference Paper
Abstract
Workers at factories or construction sites are exposed to many risks. Especially, there is an increasing tendency to occur due to collision with an internal moving means such as a forklift. There are many efforts such as attaching a sensor and a camera to a moving means, but many accidents still occur. In this paper we propose a method to prevent such accidents. First, we attach a sensor to a forklift to detect a moving object, and use a deep learning based detector to detect nearby objects from the camera output. Then, anomaly detection algorithm is used to detect a moving object even if the camera moves, and the distance between the moving object and the forklift detected through parallel distance estimation is estimated. These two results are used to calculate the risk, and when a moving object arrives within a certain range of the vehicle, the controller informs the driver of the danger signal and controls it in an emergency.
KSP Keywords
Anomaly detection algorithm, Moving Object Detection, construction site, deep learning(DL), distance estimation, nearby objects, portable camera