Design and implementation of real-time self-driving car driving using convolution neural networks and IoT
Authors: Duong Dinh Tu, Mai The Anh, Ho Sy Phuong, Le Van Chuong, Ta Hung Cuong, Nguyen Xuan Hung, Tran Huy Hoang
THE 8th ACADEMIC CONFERENCE ON NATURAL SCIENCE FOR YOUNG SCIENTISTS, MASTER AND PhD STUDENTS FROM ASEAN COUNTRIES
: 8 : 651-658
Publishing year: 8/2023
Advancements in technology have brought the possibility of fully autonomous vehicles closer to reality, making self-driving cars an increasingly popular topic of interest. This paper proposes a real-time self-driving car prototype using CNN and IoT. The car performs three tasks: lane detection with a CNN model, object detection with a YOLO model, and data transmission directly to the server. The web server is built to control and monitor the vehicle over the internet. The experimental results demonstrate that the proposed car was able to achieve a reasonable degree of accuracy in lane detection and object detection, with a very low data transmission delay to the server. Overall, this real-time self-driving car prototype represents a promising step towards the development of fully autonomous vehicles.
