Remote sensing image change detection based on slow feature analysis guided multi-level attention autoencoder

doi:10.13474/j.cnki.11-2246.2024.0913

Abstract

Abstract: Remote sensing image change detection is an important way to identify and quantify surface changes, and is one of the main applications of remote sensing technology. However, in remote sensing images obtained under different imaging conditions such as lighting and seasons, the same object may exhibit different appearances, making it difficult for change detection algorithms to accurately distinguish real ground changes. A remote sensing image change detection method based on slow feature analysis guidance multi-level attention autoencoder (SFAMAA) is proposed to address this issue. Firstly, a U-shaped convolutional autoencoder is designed and a multi-level channel attention mechanism is introduced to expand the network's receptive field while focusing on important channel information, enhancing the network's perception of global and changing information; In addition, a slow feature analysis loss function is designed to guide network training, enabling the network to effectively suppress pseudo changes caused by differences in imaging conditions. Experimental verification is conducted on the public dataset SZTAKI, and the results show that the proposed method can effectively suppress noise and pseudo changes, and has high accuracy and good robustness for remote sensing images obtained under different imaging conditions such as lighting and seasons.

Key words: autoencoder, channel attention, slow feature analysis, change detection

CLC Number:

LIU Jinling, CHEN Hongzhen, LI Chenzheng, NIE Hongbin, LI Ligang. Remote sensing image change detection based on slow feature analysis guided multi-level attention autoencoder[J]. Bulletin of Surveying and Mapping, 2024, 0(9): 67-73.

References

[1] 任秋如, 杨文忠, 汪传建, 等.遥感影像变化检测综述[J].计算机应用, 2021, 41(8):2294-2305.
[2] BOVOLO F, BRUZZONE L.A theoretical framework for unsupervised change detection based on change vector analysis in the polar domain[J].IEEE Transactions on Geoscience and Remote Sensing, 2007, 45(1):218-236.
[3] CELIK T.Unsupervised change detection in satellite images using principal component analysis and $k$-means clustering[J].IEEE Geoscience and Remote Sensing Letters, 2009, 6(4):772-776.
[4] WU Chen, DU Bo, ZHANG Liangpei.Slow feature analysis for change detection in multispectral imagery[J].IEEE Transactions on Geoscience and Remote Sensing, 2014, 52(5):2858-2874.
[5] 陈婕, 刘纪平, 徐胜华.增强边缘信息的全卷积神经网络遥感影像建筑物变化检测[J].测绘通报, 2023(6):61-67.
[6] 梁哲恒, 黎宵, 邓鹏, 等.融合多尺度特征注意力的遥感影像变化检测方法[J].测绘学报, 2022, 51(5):668-676.
[7] DU Bo, RU Lixiang, WU Chen, et al.Unsupervised deep slow feature analysis for change detection in multi-temporal remote sensing images[J].IEEE Transactions on Geoscience and Remote Sensing, 2019, 57(12):9976-9992.
[8] BERGAMASCO L, SAHA S, BOVOLO F, et al.Unsupervised change detection using convolutional-autoencoder multiresolution features[J].IEEE Transactions on Geoscience and Remote Sensing, 2022, 60:4408119.
[9] LUPPINO L T, HANSEN M A, KAMPFFMEYER M, et al.Code-aligned autoencoders for unsupervised change detection in multimodal remote sensing images[J].IEEE Transactions on Neural Networks and Learning Systems, 2022, 35:60-72.
[10] JIA Meng, ZHANG Cheng, ZHAO Zhiqiang, et al.Bipartite graph attention autoencoders for unsupervised change detection using VHR remote sensing images[J].IEEE Transactions on Geoscience and Remote Sensing, 2022, 60:5626215.
[11] OTSU N.A threshold selection method from gray-level histograms[J].IEEE Transactions on Systems, Man, and Cybernetics, 1979, 9(1):62-66.
[12] WOO S, PARK J, LEE J Y, et al.CBAM:convolutional block attention module[M]//Computer Vision - ECCV 2018.Cham:Springer International Publishing, 2018:3-19.
[13] WISKOTT L, SEJNOWSKI T J.Slow feature analysis:unsupervised learning of invariances[J].Neural Computation, 2002, 14(4):715-770.
[14] BENEDEK C, SZIRANYI T.Change detection in optical aerial images by a multilayer conditional mixed Markov model[J].IEEE Transactions on Geoscience and Remote Sensing, 2009, 47(10):3416-3430.
[15] NIELSEN A A.The regularized iteratively reweighted MAD method for change detection in multi- and hyperspectral data[J].IEEE Transactions on Image Processing:a Publication of the IEEE Signal Processing Society, 2007, 16(2):463-478.