UVid-Net: Enhanced Semantic Segmentation of UAV Aerial Videos by Embedding Temporal Information

Girisha, S and Verma, Ujwal and Pai, Manohara M.M. and Pai, Radhika M (2021) UVid-Net: Enhanced Semantic Segmentation of UAV Aerial Videos by Embedding Temporal Information. Ieee journal of selected topics in applied earth observations and remote sensing, 14. pp. 4115-4127. ISSN 1939-1404

[img] PDF
12377.pdf - Published Version
Restricted to Registered users only

Download (5MB) | Request a copy
Official URL: https://creativecommons.org/licenses/by-nc-nd/4.0/

Abstract

Semantic segmentation of aerial videos has been ex�tensively used for decision making in monitoring environmental changes, urban planning, and disaster management. The reliability of these decision support systems is dependent on the accuracy of the video semantic segmentation algorithms. The existing CNN�based video semantic segmentation methods have enhanced the image semantic segmentation methods by incorporating an addi�tional module such as LSTM or optical flow for computing temporal dynamics of the video which is a computational overhead. The proposed research work modifies the CNN architecture by incor�porating temporal information to improve the efficiency of video semantic segmentation. In this work, an enhanced encoder–decoder based CNN architecture (UVid-Net) is proposed for unmanned aerial vehicle (UAV) video semantic segmentation. The encoder of the proposed architecture embeds temporal information for tem�porally consistent labeling. The decoder is enhanced by introducing the feature-refiner module, which aids in accurate localization of the class labels. The proposed UVid-Net architecture for UAV video semantic segmentation is quantitatively evaluated on extended ManipalUAVid dataset. The performance metric mean Intersection over Union of 0.79 has been observed which is significantly greater than the other state-of-the-art algorithms. Further, the proposed work produced promising results even for the pretrained model of UVid-Net on urban street scene by fine tuning the final layer on UAV aerial videos

Item Type: Article
Uncontrolled Keywords: —Deep learning, semantic segmentation, transfer learning, U-Net, unmanned aerial vehicle (UAV) video.
Subjects: Engineering > MIT Manipal > Electronics and Communication
Engineering > MIT Manipal > Information and Communication Technology
Depositing User: MIT Library
Date Deposited: 05 Oct 2021 09:29
Last Modified: 05 Oct 2021 09:29
URI: http://eprints.manipal.edu/id/eprint/157479

Actions (login required)

View Item View Item