Dual-Domain Fusion Network Based on Wavelet Frequency Decomposition and Fuzzy Spatial Constraint for Remote Sensing Image Segmentation

被引:1
|
作者
Wei, Guangyi [1 ]
Xu, Jindong [1 ]
Yan, Weiqing [1 ]
Chong, Qianpeng [2 ]
Xing, Haihua [3 ]
Ni, Mengying [1 ]
机构
[1] Yantai Univ, Sch Comp & Control Engn, Yantai 264005, Peoples R China
[2] Beijing Normal Univ, Sch Artificial Intelligence, Beijing 100875, Peoples R China
[3] Hainan Normal Univ, Sch Informat Sci & Technol, Haikou 571158, Peoples R China
基金
中国国家自然科学基金;
关键词
remote sensing; semantic segmentation; wavelet transform; type-2; fuzzy; SEMANTIC SEGMENTATION; CHALLENGES; CLASSIFICATION; FRAMEWORK;
D O I
10.3390/rs16193594
中图分类号
X [环境科学、安全科学];
学科分类号
08 ; 0830 ;
摘要
Semantic segmentation is crucial for a wide range of downstream applications in remote sensing, aiming to classify pixels in remote sensing images (RSIs) at the semantic level. The dramatic variations in grayscale and the stacking of categories within RSIs lead to unstable inter-class variance and exacerbate the uncertainty around category boundaries. However, existing methods typically emphasize spatial information while overlooking frequency insights, making it difficult to achieve desirable results. To address these challenges, we propose a novel dual-domain fusion network that integrates both spatial and frequency features. For grayscale variations, a multi-level wavelet frequency decomposition module (MWFD) is introduced to extract and integrate multi-level frequency features to enhance the distinctiveness between spatially similar categories. To mitigate the uncertainty of boundaries, a type-2 fuzzy spatial constraint module (T2FSC) is proposed to achieve flexible higher-order fuzzy modeling to adaptively constrain the boundary features in the spatial by constructing upper and lower membership functions. Furthermore, a dual-domain feature fusion (DFF) module bridges the semantic gap between the frequency and spatial features, effectively realizes semantic alignment and feature fusion between the dual domains, which further improves the accuracy of segmentation results. We conduct comprehensive experiments and extensive ablation studies on three well-known datasets: Vaihingen, Potsdam, and GID. In these three datasets, our method achieved 74.56%, 73.60%, and 81.01% mIoU, respectively. Quantitative and qualitative results demonstrate that the proposed method significantly outperforms state-of-the-art methods, achieving an excellent balance between segmentation accuracy and computational overhead.
引用
收藏
页数:24
相关论文
共 50 条
  • [21] Wavelet Frequency Channel Attention on Remote Sensing Image Segmentation
    Su, Yu-Chen
    Liu, Tsung-Jung
    Liu, Kuan-Hsien
    2022 IEEE INTERNATIONAL CONFERENCE ON CONSUMER ELECTRONICS - TAIWAN, IEEE ICCE-TW 2022, 2022, : 285 - 286
  • [22] Remote Sensing Image Dehazing Based on Dual Attention Parallelism and Frequency Domain Selection Network
    Su, Hang
    Liu, Lina
    Jeon, Gwanggil
    Wang, Zenghui
    Guo, Tiancun
    Gao, Mingliang
    IEEE TRANSACTIONS ON CONSUMER ELECTRONICS, 2024, 70 (03) : 5300 - 5311
  • [23] The use of wavelet high frequency substitution fusion to increase remote sensing image spatial resolution
    Bagan, H
    Ma, JW
    Li, QQ
    Liu, ZL
    Han, XZ
    THIRD INTERNATIONAL SYMPOSIUM ON MULTISPECTRAL IMAGE PROCESSING AND PATTERN RECOGNITION, PTS 1 AND 2, 2003, 5286 : 376 - 379
  • [24] Modified multiscale fusion combing wavelet-domain HMT model for remote sensing image segmentation
    Sun, Wen-Feng
    Sun, Qiang
    Jiao, Li-Cheng
    Hongwai yu Jiguang Gongcheng/Infrared and Laser Engineering, 2004, 33 (05): : 528 - 532
  • [25] Remote Sensing Image Fusion Based on Wavelet Techniques
    Qi, Yuan
    2009 ASIA-PACIFIC CONFERENCE ON INFORMATION PROCESSING (APCIP 2009), VOL 2, PROCEEDINGS, 2009, : 84 - 87
  • [26] Remote sensing image fusion based on IBSM and wavelet
    Chen, Xiao-Mei
    Ni, Guo-Qiang
    Li, Yong-Liang
    Beijing Ligong Daxue Xuebao/Transaction of Beijing Institute of Technology, 2008, 28 (SUPPL.): : 90 - 93
  • [27] Dual-domain sampling and feature-domain optimization network for image compressive sensing
    Xiang, Xinxin
    Tong, Fenghua
    Zhao, Dawei
    Li, Xin
    Yang, Shumian
    ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE, 2024, 137
  • [28] Remote sensing image segmentation based on the fuzzy deep convolutional neural network
    Zhao, Tianyu
    Xu, Jindong
    Chen, Rui
    Ma, Xiangyue
    INTERNATIONAL JOURNAL OF REMOTE SENSING, 2021, 42 (16) : 6267 - 6286
  • [29] LIGHTWEIGHT DUAL-DOMAIN NETWORK FOR REAL-TIME MEDICAL IMAGE SEGMENTATION
    Liu, Libo
    Fan, Xinxin
    Zhang, Xiaodong
    Hu, Qingmao
    2022 IEEE INTERNATIONAL CONFERENCE ON IMAGE PROCESSING, ICIP, 2022, : 396 - 400
  • [30] Decomposition-Based Unsupervised Domain Adaptation for Remote Sensing Image Semantic Segmentation
    Ma, Xianping
    Zhang, Xiaokang
    Ding, Xingchen
    Pun, Man-On
    Ma, Siwei
    IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2024, 62