Res50-SimAM-ASPP-Unet: A Semantic Segmentation Model for High-Resolution Remote Sensing Images

High-resolution remote sensing images contain intricate details and complex backgrounds, presenting challenges for traditional segmentation methods, which often struggle with accurate classification and contextual understanding. To address these issues, this study introduces the Res50-SimAM-ASPP-Une...

Full description

Saved in:
Bibliographic Details
Main Authors: Jiajing Cai, Jinmei Shi, Yu-Beng Leau, Shangyu Meng, Xiuyan Zheng, Jinghe Zhou
Format: Article
Language:English
Published: IEEE 2024-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10804789/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846113902301020160
author Jiajing Cai
Jinmei Shi
Yu-Beng Leau
Shangyu Meng
Xiuyan Zheng
Jinghe Zhou
author_facet Jiajing Cai
Jinmei Shi
Yu-Beng Leau
Shangyu Meng
Xiuyan Zheng
Jinghe Zhou
author_sort Jiajing Cai
collection DOAJ
description High-resolution remote sensing images contain intricate details and complex backgrounds, presenting challenges for traditional segmentation methods, which often struggle with accurate classification and contextual understanding. To address these issues, this study introduces the Res50-SimAM-ASPP-Unet model, a semantic segmentation approach for high-resolution remote sensing image processing tasks. The model integrates ResNet50 as the encoding layer of Unet for robust feature extraction, adds the SimAM attention mechanism to selectively enhance relevant details, and incorporates the ASPP module in the decoding layer to capture multi-scale contextual information. The methodology part analyzes the common ResNet model, the attention mechanism module, and the multi-scale feature extraction module, respectively, and then designs experiments to show the necessity and optimal position of adding Res50, SimAM, and ASPP. Comparative experiments on the LandCover.ai dataset demonstrate that the proposed model significantly outperforms common semantic segmentation networks, achieving a MIoU of 81.1%, MPA of 88.2%, Accuracy of 95.1%, Precision of 92.65%, and an F1 score of 90.45%. These results highlight the model’s effectiveness in delivering high accuracy and adaptability across diverse remote sensing environments, establishing it as a valuable tool for applications requiring precise and scalable image segmentation.
format Article
id doaj-art-ddf8fe5e7a2c4c5eaa955b24d73d2592
institution Kabale University
issn 2169-3536
language English
publishDate 2024-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-ddf8fe5e7a2c4c5eaa955b24d73d25922024-12-21T00:01:03ZengIEEEIEEE Access2169-35362024-01-011219230119231610.1109/ACCESS.2024.351926010804789Res50-SimAM-ASPP-Unet: A Semantic Segmentation Model for High-Resolution Remote Sensing ImagesJiajing Cai0https://orcid.org/0000-0002-4751-3028Jinmei Shi1https://orcid.org/0000-0002-9021-5464Yu-Beng Leau2https://orcid.org/0000-0002-5386-2734Shangyu Meng3https://orcid.org/0009-0001-1836-289XXiuyan Zheng4https://orcid.org/0009-0007-1034-9950Jinghe Zhou5https://orcid.org/0000-0002-6563-1263College of Information Engineering, Hainan Vocational University of Science and Technology, Haikou, ChinaCollege of Information Engineering, Hainan Vocational University of Science and Technology, Haikou, ChinaFaculty of Computing and Informatics, Universiti Malaysia Sabah, Kota Kinabalu, MalaysiaSchool of Information Science and Technology, Universiti Kebangsaan Malaysia (UKM), Bangi, Selangor, MalaysiaDepartment of Information Engineering, Guangzhou Baiyun Industrial and Commercial Technician College, Guangzhou, ChinaCollege of Information Engineering, Hainan Vocational University of Science and Technology, Haikou, ChinaHigh-resolution remote sensing images contain intricate details and complex backgrounds, presenting challenges for traditional segmentation methods, which often struggle with accurate classification and contextual understanding. To address these issues, this study introduces the Res50-SimAM-ASPP-Unet model, a semantic segmentation approach for high-resolution remote sensing image processing tasks. The model integrates ResNet50 as the encoding layer of Unet for robust feature extraction, adds the SimAM attention mechanism to selectively enhance relevant details, and incorporates the ASPP module in the decoding layer to capture multi-scale contextual information. The methodology part analyzes the common ResNet model, the attention mechanism module, and the multi-scale feature extraction module, respectively, and then designs experiments to show the necessity and optimal position of adding Res50, SimAM, and ASPP. Comparative experiments on the LandCover.ai dataset demonstrate that the proposed model significantly outperforms common semantic segmentation networks, achieving a MIoU of 81.1%, MPA of 88.2%, Accuracy of 95.1%, Precision of 92.65%, and an F1 score of 90.45%. These results highlight the model’s effectiveness in delivering high accuracy and adaptability across diverse remote sensing environments, establishing it as a valuable tool for applications requiring precise and scalable image segmentation.https://ieeexplore.ieee.org/document/10804789/Segmentation of high-resolution remote sensing imagesmulti-scale void space pyramid pool ASPP moduleattention mechanism SimAM moduleRes50-SimAM-ASPP-Unet
spellingShingle Jiajing Cai
Jinmei Shi
Yu-Beng Leau
Shangyu Meng
Xiuyan Zheng
Jinghe Zhou
Res50-SimAM-ASPP-Unet: A Semantic Segmentation Model for High-Resolution Remote Sensing Images
IEEE Access
Segmentation of high-resolution remote sensing images
multi-scale void space pyramid pool ASPP module
attention mechanism SimAM module
Res50-SimAM-ASPP-Unet
title Res50-SimAM-ASPP-Unet: A Semantic Segmentation Model for High-Resolution Remote Sensing Images
title_full Res50-SimAM-ASPP-Unet: A Semantic Segmentation Model for High-Resolution Remote Sensing Images
title_fullStr Res50-SimAM-ASPP-Unet: A Semantic Segmentation Model for High-Resolution Remote Sensing Images
title_full_unstemmed Res50-SimAM-ASPP-Unet: A Semantic Segmentation Model for High-Resolution Remote Sensing Images
title_short Res50-SimAM-ASPP-Unet: A Semantic Segmentation Model for High-Resolution Remote Sensing Images
title_sort res50 simam aspp unet a semantic segmentation model for high resolution remote sensing images
topic Segmentation of high-resolution remote sensing images
multi-scale void space pyramid pool ASPP module
attention mechanism SimAM module
Res50-SimAM-ASPP-Unet
url https://ieeexplore.ieee.org/document/10804789/
work_keys_str_mv AT jiajingcai res50simamasppunetasemanticsegmentationmodelforhighresolutionremotesensingimages
AT jinmeishi res50simamasppunetasemanticsegmentationmodelforhighresolutionremotesensingimages
AT yubengleau res50simamasppunetasemanticsegmentationmodelforhighresolutionremotesensingimages
AT shangyumeng res50simamasppunetasemanticsegmentationmodelforhighresolutionremotesensingimages
AT xiuyanzheng res50simamasppunetasemanticsegmentationmodelforhighresolutionremotesensingimages
AT jinghezhou res50simamasppunetasemanticsegmentationmodelforhighresolutionremotesensingimages