ETRI-Knowledge Sharing Plaform

KOREAN
논문 검색
Type SCI
Year ~ Keyword

Detail

Conference Paper Analysis of Operation Algorithm for Intelligent Multi-Surveillance Image Sensors
Cited 0 time in scopus Share share facebook twitter linkedin kakaostory
Authors
Young-il Kim, Tae-Chong Lee, Yonghyun Yeun, Jae-Young Lee, Sungeun Jin, Khishigjargal Gonchigsumlaa
Issue Date
2023-10
Citation
Asia Pacific Conference on Wireless and Mobile (APWiMob) 2023, pp.1-7
Language
English
Type
Conference Paper
DOI
https://dx.doi.org/10.1109/APWiMob59963.2023.10365623
Abstract
In order to detect an intruder using image deep learning in an intruder protection area, it is important to install and operate a small number of image sensor cameras efficiently. When using multiple image sensor cameras, it is necessary not only to apply handover (handoff) algorithms between cameras based on the movement of intruders but also to develop PTZ (Pan-Tilt-Zoom) control algorithms to monitor a wide area with a small number of image sensor cameras. In this paper, we analyze the average distance from intruders to the boundaries of the camera's imaging area using mathematical and simulation model, which is necessary for the PTZ control and handover algorithms of image sensor by combining the velocity and location of intruder. Additionally, the average dwell time of the intruder within the camera's monitoring area is analyzed, considering the intruder's detection location and movement speed, alongside an assessment of the handover probability. Additionally, we mathematically suggested the handover occurrence probability according to the size of the handover zone during inter-camera handover operations. Also, the probabilities of intruder handover failure and intruder detection failure are analyzed through simulations.
KSP Keywords
Average Dwell Time, Average distance, Handover algorithm, Handover failure, Handover probability, Image Sensor, Movement speed, PTZ control, Pan-Tilt-Zoom, Simulation Model, Wide area