Open AccessResearch on Assisted Driving Technology based on Improved YOLOv3
Author(s) -
Long Zhao,
Xiaoye Liu,
Qiang Wang,
Honglong Chen
Publication year - 2020
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/719/1/012062
Subject(s) - convolutional neural network , deep learning , computer science , artificial intelligence , function (biology) , artificial neural network , china , geography , evolutionary biology , biology , archaeology
In the era of automobile popularization, China and the developed countries such as Europe and the United States are facing the same problem of high car accidents. This paper uses deep learning technology to detect and identify lane lines, traffic lights, vehicles and pedestrians during driver driving. Improve the efficiency and accuracy of convolutional neural network training through migration learning and data enhancement techniques. Based on the current advanced YOLOv3 network, we have improved the network structure and loss function. The KITTI dataset has achieved the highest 2D target recognition accuracy, and the target recognition speed is higher than 36 frames/sec. The price of the assisted driving equipment we developed does not exceed RMB 5, 000, which has a good market prospect.