Skip navigation
Please use this identifier to cite or link to this item: https://libeldoc.bsuir.by/handle/123456789/54316
Title: Comparative Analysis of Semantic Segmentation Methods for Satellite Images Segmentation
Authors: Qing Bu
Wei Wan
Savitskaya, E.
Keywords: материалы конференций;semantic segmentation;image segmentation;urban scenes;deep neural network;U-Net;CNN-based semantic segmentation;transformers
Issue Date: 2023
Publisher: BSU
Citation: Qing Bu. Comparative Analysis of Semantic Segmentation Methods for Satellite Images Segmentation / Qing Bu, Wei Wan, E. Savitskaya // Pattern Recognition and Information Processing (PRIP'2023) = Распознавание образов и обработка информации (2023) : Proceedings of the 16th International Conference, October 17–19, 2023, Minsk, Belarus / United Institute of Informatics Problems of the National Academy of Sciences of Belarus. – Minsk, 2023. – P. 332–337.
Abstract: This paper proposes a comparative analysis of different automatic semantic segmentation methods for satellite images segmentation on the Semantic Drone Dataset with 23 classes (paved-area, dirt, grass, gravel, water, rocks, pool, vegetation, roof, wall, window, door, fence, fence-pole, person, dog, car, bicycle, tree, bald-tree, ar-marker, obstacle, conflicting). We compare such models as U-net, U-net++, FPN, PAN, DeepLabV3, DeepLabV3+ and Transformer architecture model - SegFormer.
URI: https://libeldoc.bsuir.by/handle/123456789/54316
Appears in Collections:Pattern Recognition and Information Processing (PRIP'2023) = Распознавание образов и обработка информации (2023)

Files in This Item:
File Description SizeFormat 
Qing_Bu_Comparative.pdf1.4 MBAdobe PDFView/Open
Show full item record Google Scholar

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.