Attention Mask R-CNN for Ship Detection and Segmentation From Remote Sensing Images

In recent years, ship detection in satellite remote sensing images has become an important research topic. Most existing methods detect ships by using a rectangular bounding box but do not perform segmentation down to the pixel level. This paper proposes a ship detection and segmentation method base...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Xuan Nie, Mengyang Duan, Haoxuan Ding, Bingliang Hu, Edward K. Wong
Formato:	article
Lenguaje:	EN
Publicado:	IEEE 2020
Materias:	Computer vision object detection object segmentation remote sensing Electrical engineering. Electronics. Nuclear engineering TK1-9971
Acceso en línea:	https://doaj.org/article/ea643b7985fb4a7bb90c8ca63e24c9e2
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:ea643b7985fb4a7bb90c8ca63e24c9e2
record_format	dspace
spelling	oai:doaj.org-article:ea643b7985fb4a7bb90c8ca63e24c9e22021-11-19T00:03:24ZAttention Mask R-CNN for Ship Detection and Segmentation From Remote Sensing Images2169-353610.1109/ACCESS.2020.2964540https://doaj.org/article/ea643b7985fb4a7bb90c8ca63e24c9e22020-01-01T00:00:00Zhttps://ieeexplore.ieee.org/document/8951182/https://doaj.org/toc/2169-3536In recent years, ship detection in satellite remote sensing images has become an important research topic. Most existing methods detect ships by using a rectangular bounding box but do not perform segmentation down to the pixel level. This paper proposes a ship detection and segmentation method based on an improved Mask R-CNN model. Our proposed method can accurately detect and segment ships at the pixel level. By adding a bottom-up structure to the FPN structure of Mask R-CNN, the path between the lower layers and the topmost layer is shortened, allowing the lower layer features to be more effectively utilized at the top layer. In the bottom-up structure, we use channel-wise attention to assign weights in each channel and use the spatial attention mechanism to assign a corresponding weight at each pixel in the feature maps. This allows the feature maps to respond better to the target’s features. Using our method, the detection and segmentation mAPs increased from 70.6% and 62.0% to 76.1% and 65.8%, respectively.Xuan NieMengyang DuanHaoxuan DingBingliang HuEdward K. WongIEEEarticleComputer visionobject detectionobject segmentationremote sensingElectrical engineering. Electronics. Nuclear engineeringTK1-9971ENIEEE Access, Vol 8, Pp 9325-9334 (2020)
institution	DOAJ
collection	DOAJ
language	EN
topic	Computer vision object detection object segmentation remote sensing Electrical engineering. Electronics. Nuclear engineering TK1-9971
spellingShingle	Computer vision object detection object segmentation remote sensing Electrical engineering. Electronics. Nuclear engineering TK1-9971 Xuan Nie Mengyang Duan Haoxuan Ding Bingliang Hu Edward K. Wong Attention Mask R-CNN for Ship Detection and Segmentation From Remote Sensing Images
description	In recent years, ship detection in satellite remote sensing images has become an important research topic. Most existing methods detect ships by using a rectangular bounding box but do not perform segmentation down to the pixel level. This paper proposes a ship detection and segmentation method based on an improved Mask R-CNN model. Our proposed method can accurately detect and segment ships at the pixel level. By adding a bottom-up structure to the FPN structure of Mask R-CNN, the path between the lower layers and the topmost layer is shortened, allowing the lower layer features to be more effectively utilized at the top layer. In the bottom-up structure, we use channel-wise attention to assign weights in each channel and use the spatial attention mechanism to assign a corresponding weight at each pixel in the feature maps. This allows the feature maps to respond better to the target’s features. Using our method, the detection and segmentation mAPs increased from 70.6% and 62.0% to 76.1% and 65.8%, respectively.
format	article
author	Xuan Nie Mengyang Duan Haoxuan Ding Bingliang Hu Edward K. Wong
author_facet	Xuan Nie Mengyang Duan Haoxuan Ding Bingliang Hu Edward K. Wong
author_sort	Xuan Nie
title	Attention Mask R-CNN for Ship Detection and Segmentation From Remote Sensing Images
title_short	Attention Mask R-CNN for Ship Detection and Segmentation From Remote Sensing Images
title_full	Attention Mask R-CNN for Ship Detection and Segmentation From Remote Sensing Images
title_fullStr	Attention Mask R-CNN for Ship Detection and Segmentation From Remote Sensing Images
title_full_unstemmed	Attention Mask R-CNN for Ship Detection and Segmentation From Remote Sensing Images
title_sort	attention mask r-cnn for ship detection and segmentation from remote sensing images
publisher	IEEE
publishDate	2020
url	https://doaj.org/article/ea643b7985fb4a7bb90c8ca63e24c9e2
work_keys_str_mv	AT xuannie attentionmaskrcnnforshipdetectionandsegmentationfromremotesensingimages AT mengyangduan attentionmaskrcnnforshipdetectionandsegmentationfromremotesensingimages AT haoxuanding attentionmaskrcnnforshipdetectionandsegmentationfromremotesensingimages AT binglianghu attentionmaskrcnnforshipdetectionandsegmentationfromremotesensingimages AT edwardkwong attentionmaskrcnnforshipdetectionandsegmentationfromremotesensingimages
_version_	1718420702556585984

Attention Mask R-CNN for Ship Detection and Segmentation From Remote Sensing Images

Ejemplares similares