Date of Award
5-2023
Document Type
Thesis
Degree Name
Master of Science
Degree Discipline
Electrical Engineering
Abstract
Recently, computer vision has been promoted by deep learning techniques significantly, where supervised deep learning outperformed other methods such as in image segmentation. However, a large amount of annotated/labeled data is needed for training supervised deep learning models, while such big annotated data is typically unavailable in practice such as in satellite imagery analytics. In order to address this challenge, a novel ensemble unsupervised semantic segmentation method was proposed for image segmentation on satellite images. Specifically, an unsupervised semantic segmentation model was employed to implement foreground- background separation and then be placed within an ensemble model to increase the prediction accuracy further. Experimental results demonstrated that the proposed method outperformed baseline models such as k-means on a satellite image benchmark, the XView2 dataset. The proposed approach provides a promising solution to semantic segmentation in images that will benefit many mission-critical applications such as disaster relief using satellite imagery analytics.
Index Terms - Convolution neural network (CNNs); deep learning; ensemble model; image segmentation; overhead imagery; unsupervised learning
Committee Chair/Advisor
Xiangfang Li
Committee Member
Lijun Qian
Committee Member
Xishuang Dong
Committee Member
Pamela Obiomon
Publisher
Prairie View A&M University
Rights
© 2021 Prairie View A & M University
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
Date of Digitization
6/16/2023
Contributing Institution
John B Coleman Library
City of Publication
Prairie View
MIME Type
Application/PDF
Recommended Citation
Tarry, J. (2023). Ensemble Unsupervised Semantic Segmentation For Foreground-Background Separation On Satellite Image. Retrieved from https://digitalcommons.pvamu.edu/pvamu-theses/1518