Study on Electric Vehicle Charging Socket Detection Using YOLOv8s Model

EasyChair Preprint no. 12250

Abstract

This paper introduces the utilization of the latest small You Only Look Once version 8 – YOLOv8s convolutional neural network in an automatic electric vehicle charging application study. The employment of a deep learning based object detector is a novel and significant aspect in robotic applications, since it is both, the initial and the fundamental step in a series of robotic operations, where the intent is to detect and locate the charging socket on the vehicle’s body surface. The aim was to use a renowned and reliable object detector to ensure the reliable and smooth functioning of the deployed robotic vision system in an industrial environment. The experiments demonstrated, that the deployed YOLOv8s model detects the charging socket successfully under various image capturing conditions, with a detection rate of 97.23%.

Keyphrases: automotive applications, Electric vehicle charging socket, image processing, object detection, Robotic applications, YOLOv8s

@Booklet{EasyChair:12250,
  author = {Vladimir Tadic and Akos Odry and Zoltan Vizvari and Zoltan Kiraly and Imre Felde and Peter Odry},
  title = {Study on Electric Vehicle Charging Socket Detection Using YOLOv8s Model},
  howpublished = {EasyChair Preprint no. 12250},

  year = {EasyChair, 2024}}