Lane Type Classification & Distance Measurement System for Autonomous Vehicle Otonom Araç için Şerit Tipi Siniflandirma & Mesafe Ölçüm Sistemi


Özen S., Kaya U., Semiz A., Yalçin M. F., Çelebi A. T.

31st IEEE Conference on Signal Processing and Communications Applications, SIU 2023, İstanbul, Turkey, 5 - 08 July 2023 identifier

  • Publication Type: Conference Paper / Full Text
  • Doi Number: 10.1109/siu59756.2023.10224048
  • City: İstanbul
  • Country: Turkey
  • Keywords: custom dataset, deep learning, image processing, lane distance measurement, lane type classification
  • Istanbul Technical University Affiliated: No

Abstract

In this paper, lane type classification and lane distance measurement system are proposed for autonomous vehicles. In the proposed system, the perspective transformation method is applied to the image taken from the vehicle camera, so that a bird's-eye view is obtained and the parts with stripes are cropped from the image, and then a multi-class classification model is implemented using neural network-based architectures to determine the stripe type. In addition, with the proposed distance measurement system, the distance of the vehicle to the right and left lanes is calculated during autonomous driving, thus ensuring that the vehicle can drive autonomously on the lane. For this study, our own data set has been created.