Comparative experimental study of bridge detection methods based on deep learning

Lu Ding; Zige An; Yuanyuan Liu; Xiaochong Tong

doi:10.1117/12.2682655

1 June 2023 Comparative experimental study of bridge detection methods based on deep learning

Lu Ding, Zige An, Yuanyuan Liu, Xiaochong Tong

Proceedings Volume 12710, International Conference on Remote Sensing, Surveying, and Mapping (RSSM 2023); 127100M (2023) https://doi.org/10.1117/12.2682655
Event: International Conference on Remote Sensing, Surveying, and Mapping (RSSM 2023), 2023, Changsha, China

Abstract

As a kind of transportation hub, bridge has important social and military value. This paper constructs bridge data sets of multi-source satellite remote sensing images, and realize the automatic bridge detection based on deep learning method to solve the problem that traditional method has bad results and poor stability. The effect of Faster RCNN and YOLO-v3 in remote sensing image detection of different types of bridges are mainly studied, and the advantages and disadvantages of two typical deep learning methods in bridge detection are analyzed through experiments. Through the experiment, the two methods can achieve the automatic bridge detection and the accuracy is all over 90%. In terms of detection accuracy, YOLO-v3 is higher than Faster RCNN(AP50), but its detection stability is lower than Faster RCNN. In addition, the training time and single image detection time of YOLO-v3 are both better than that of Faster RCNN.

Citation Download Citation

Lu Ding, Zige An, Yuanyuan Liu, and Xiaochong Tong "Comparative experimental study of bridge detection methods based on deep learning", Proc. SPIE 12710, International Conference on Remote Sensing, Surveying, and Mapping (RSSM 2023), 127100M (1 June 2023); https://doi.org/10.1117/12.2682655

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available