Sains Malaysiana 55(2)(2026): 209-219

http://doi.org/10.17576/jsm-2026-5502-03

 

A Review of CNN-Based Typical Urban Land Cover Segmentation Techniques in Multispectral Remote Sensing Imagery

(Suatu Ulasan Teknik Segmentasi Litupan Tanah Bandar Tipikal Berasaskan CNN dalam Imej Penderiaan Jauh Multispektral)

 

ZHAO HAIMENG^1,2, RAIHANI MOHAMED^1,* & NG SENG BENG¹

 

¹Faculty of Computer Science and Information Technology, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia

²College of Artificial Intelligence, Guilin University of Aerospace Technology, Guilin, Guangxi, 541004, China

 

Received: 20 June 2025/Accepted: 28 January 2026

 

Abstract

Compared with visible-light remote sensing, multispectral remote sensing provides multi-band land surface information and enhances spectral separability through data fusion, thereby enabling more accurate surface representation. However, spectral redundancy, resolution discrepancies, and highly complex urban environments impose greater challenges on existing methods. Deep learning approaches based on convolutional neural network (CNN) offer superior capabilities in extracting and integrating multispectral features, enabling more accurate urban land cover segmentation. This review focuses on pixel-level urban land cover segmentation and systematically summarizes recent advances in deep learning for multispectral remote sensing. First, we emphasize that the rich spectral information and spatial complementarity of multispectral data effectively enhance segmentation performance and alleviate ambiguities caused by the ‘same spectrum-different objects’ and ‘same object-different spectra’. Second, we review 19 publicly available multispectral datasets, highlighting differences in spectral bands, spatial resolution, and application scenarios, and summarize a standardized preprocessing pipeline including radiometric calibration, geometric correction, band normalization, and spectral dimensionality reduction to support reproducibility. Third, we discuss representative spectral-spatial feature extraction and cross-scale context modeling strategies, covering dilated convolution, 3D-2D hybrid structures, dual-branch architectures, and multi-scale enhancement modules. Extensive comparative experiments on ISPRS Potsdam and GID datasets further demonstrate the applicability and performance differences of representative models. Finally, future research trends and directions are discussed, encompassing multi-temporal and multi-scale temporal learning, cross-modal fusion, and the lightweight design of complex models.

Keywords: Convolutional neural network (CNN); multispectral features; remote sensing data; semantic segmentation; surface feature extraction

 

Abstrak

Dibandingkan dengan penderiaan jauh cahaya tampak, penderiaan jauh multispektral menyediakan maklumat permukaan tanah pelbagai jalur dan meningkatkan kebolehbezaan spektral melalui penggabungan data, seterusnya membolehkan perwakilan permukaan yang lebih tepat. Walau bagaimanapun, pertindihan spektral, perbezaan resolusi dan persekitaran bandar yang sangat kompleks menimbulkan cabaran lebih besar terhadap kaedah sedia ada. Pendekatan pembelajaran mendalam berasaskan rangkaian neural konvolusi (CNN) menawarkan keupayaan unggul dalam mengekstrak dan mengintegrasikan ciri multispektral, membolehkan pengasingan litupan tanah bandar yang lebih tepat. Ulasan ini memberi tumpuan pada pengasingan liputan tanah bandar per tahap piksel dan secara sistematik merumuskan kemajuan terkini dalam pembelajaran mendalam untuk penderiaan jauh multispektral. Pertama, kami menekankan bahawa maklumat spektral yang kaya dan pelengkap reruang data multispektral berkesan meningkatkan prestasi pengasingan dan mengurangkan kekeliruan akibat ‘spektrum sama-objek berbeza’ dan ‘objek sama-spektrum berbeza’. Kedua, kami mengkaji 19 set data multispektral yang tersedia secara awam, menyoroti perbezaan dalam jalur spektral, resolusi spasial dan senario aplikasi, serta merumuskan saluran prapemprosesan piawai termasuk kalibrasi radiometrik, pembetulan geometri, normalisasi jalur dan pengurangan dimensi spektral untuk menyokong kebolehulangan. Ketiga, kami membincangkan strategi pengekstrakan ciri spektral-reruang dan permodelan konteks silang-skala, merangkumi konvolusi dilasi, struktur hibrid 3D-2D, seni bina dwi-cabang dan modul peningkatan multi-skala. Uji kaji perbandingan luas pada dataset ISPRS Potsdam dan GID seterusnya menunjukkan keberkesanan dan perbezaan prestasi model wakil. Akhirnya, tren dan arah penyelidikan masa depan dibincangkan, termasuk pembelajaran temporal berbilang-skala dan berbilang-masa, penggabungan lintas-mod serta reka bentuk ringan bagi model kompleks.

Kata kunci: Ciri multispektral; data penderiaan jauh; pengekstrakan ciri permukaan; pengelasan semantik; rangkaian neural konvolusi (CNN)

 

REFERENCES

Alhassan, V., Henry, C., Ramanna, S. & Storie, C. 2020. A deep learning framework for land-use/land-cover mapping and analysis using multispectral satellite imagery. Neural Computing and Applications 32: 8529-8544.

Anandakrishnan, J., Sundaram, V.M. & Paneer, P. 2025. STA-AgriNet: A spatio-temporal attention framework for crop type mapping from fused multi-sensor multi-temporal SITS. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 18: 1817-1826.

Bishoff, E., Godfrey, C., McKay, M. & Byler, E. 2023. Quantifying the robustness of deep multispectral segmentation models against natural perturbations and data poisoning. In Algorithms, Technologies, and Applications for Multispectral and Hyperspectral Imaging XXIX, SPIE, 12519: 200-213.

Buttar, P.K. & Sachan, M.K. 2024. Land cover segmentation using 3D FCN-based architecture with coordinate attention. IEEE Geoscience and Remote Sensing Letters 21: 2502905.

Chen, L.C., Papandreou, G., Schroff, F. & Adam, H. 2017. Rethinking atrous convolution for semantic image segmentation.arXiv preprint arXiv:1706.05587.

Ding, L., Hong, D., Zhao, M., Chen, H., Li, C., Deng, J., Yokoya, N., Bruzzone, L. & Chanussot, J. 2025. A survey of sample-efficient deep learning for change detection in remote sensing: Tasks, strategies, and challenges. IEEE Geoscience and Remote Sensing Magazine 13(3): 164-189.

Du, Y., Sheng, Q., Zhang, W., Zhu, C., Li, J. & Wang, B. 2023. From local context-aware to non-local: A road extraction network via guidance of multi-spectral image. ISPRS Journal of Photogrammetry and Remote Sensing 203: 230-245.

Gui, Y., Li, W., Xia, X.G., Tao, R. & Yue, A. 2022. Infrared attention network for woodland segmentation using multispectral satellite images. IEEE Transactions on Geoscience and Remote Sensing 60: 5627214.

Han, Z., Tian, Q., Tian, J., Zhao, T., Xu, C. & Zhou, Q. 2025. Estimation of fractional cover based on NDVI-VISI response space using visible-near infrared satellite imagery. International Journal of Applied Earth Observation and Geoinformation 137: 104432.

He, K., Chen, X., Xie, S., Li, Y., Dollár, P. & Girshick, R. 2022. Masked autoencoders are scalable vision learners. IEEE/CVF Conf. on Computer Vision and Pattern Recognition https://doi.org/10.1109/CVPR52688.2022.01553

Hong, D., Zhang, B., Li, H., Li, Y., Yao, J., Li, C., Werner, M., Chanussot, J., Zipf, A. & Zhu, X.X. 2023. Cross-city matters: A multimodal remote sensing benchmark dataset for cross-city semantic segmentation using high-resolution domain adaptation networks. Remote Sensing of Environment 299: 113856.

Jia, J., Song, J., Kong, Q., Yang, H., Teng, Y. & Song, X. 2023. Multi-attention-based semantic segmentation network for land cover remote sensing images. Electronics 12(6): 1347.

Li, J., Cai, Y., Li, Q., Kou, M. & Zhang, T. 2024. A review of remote sensing image segmentation by deep learning methods. International Journal of Digital Earth 17: 2328827.

Li, R., Zheng, S., Zhang, C., Duan, C., Wang, L. & Atkinson, P.M. 2021. ABCNet: Attentive bilateral contextual network for efficient semantic segmentation of fine-resolution remotely sensed imagery. ISPRS Journal of Photogrammetry and Remote Sensing 181: 84-98.

Lin, L., Liu, L., Liu, M., Zhang, Q., Feng, M., Khalil, Y.S. & Yin, F. 2024. DEDNet: Dual-Encoder DeeplabV3+ network for rock glacier recognition based on multispectral remote sensing image. Remote Sensing 16(14): 2603.

Mo, W., Tan, Y., Zhou, Y., Zhi, Y., Cai, Y. & Ma, W. 2023. Multispectral remote sensing image change detection based on twin neural networks. Electronics 12(18): 3766.

Muhtar, D., Zhang, X. & Xiao, P. 2022. Index your position: A novel self-supervised learning method for remote sensing images semantic segmentation. IEEE Transactions on Geoscience and Remote Sensing 60: 4411511.

Nagaraj, R. & Kumar, L.S. 2024. Extraction of surface water bodies using optical remote sensing images: A review. Earth Science Informatics 17(2): 893-956.

Ramos, L. & Sappa, A.D. 2024. Multispectral semantic segmentation for land cover classification: An overview. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 17: 14295-14336.

Saralioglu, E. & Gungor, O. 2022. Semantic segmentation of land cover from high resolution multispectral satellite images by spectral-spatial convolutional neural network. Geocarto International 37: 657-677.

Shen, X., Weng, L., Xia, M. & Others. 2022. Multi-scale feature aggregation network for semantic segmentation of land cover. Remote Sensing 14: 6156.

Sun, J., Yin, M., Wang, Z., Xie, T. & Bei, S. 2024. Multispectral object detection based on multilevel feature fusion and dual feature modulation. Electronics 13(2): 443.

Tao, C., Meng, Y., Li, J., Yang, B., Hu, F., Li, Y., Cui, C. & Zhang, W. 2022. MSNet: Multispectral semantic segmentation network for remote sensing images. GIScience & Remote Sensing 59(1): 1177-1198.

Thisanke, H., Deshan, C., Chamith, K., Seneviratne, S., Vidanaarachchi, R. & Herath, D. 2023. Semantic segmentation using vision transformers: A survey. Engineering Applications of Artificial Intelligence 126(Part A): 106669.

Tong, Z., Li, Y., Zhang, J., He, L. & Gong, Y. 2023. MSFANet: Multiscale fusion attention network for road segmentation of multispectral remote sensing data. Remote Sensing 15(8): 1978.

Ulku, I. 2024. ContexNestedU-Net: Efficient context-aware semantic segmentation architecture for precision agriculture applications based on multispectral remote sensing imagery. Traitement du Signal 41(5): 2425-2436.

Wang, L., Li, R., Zhang, C., Fang, S., Duan, C., Meng, X. & Atkinson, P.M. 2022. UNetFormer: A UNet-like transformer for efficient semantic segmentation of remote sensing urban scene imagery. ISPRS Journal of Photogrammetry and Remote Sensing 190: 196-214.

Wang, L., Li, R., Wang, D., Duan, C., Wang, T. & Meng, X. 2021. Transformer meets convolution: A bilateral awareness network for semantic segmentation of very fine resolution urban scene images. Remote Sensing 13(16): 3065.

Wang, Q., Hu, C., Wang, H., Wang, R., Xie, Y. & Zhao, Y. 2024. Semantic segmentation of urban land classes using a multi-scale dataset. International Journal of Remote Sensing 45(2): 653-675.

Wu, X., Wang, P., Gong, Y., Zhang, Y., Wang, Q., Li, Y., Guo, J. & Han, S. 2024. Construction and application of dynamic threshold model for agricultural drought grades based on near-infrared and short-wave infrared bands for spring maize. Remote Sensing 16(17): 3260.

Xue, Z., Yang, G., Yu, X., Yu, A., Guo, Y., Liu, B. & Zhou, J. 2025. Multimodal self-supervised learning for remote sensing data land cover classification. Pattern Recognition 157: 110959.

Yan, Q., Zhang, S., Chen, X. & Zheng, Z. 2025. Multiscale superpixel depth feature extraction for hyperspectral image classification. Scientific Reports 15(1): 13529.

Yu, A., Quan, Y., Yu, R., Guo, W., Wang, X., Hong, D., Zhang, H., Chen, J., Hu, Q. & He, P. 2023. Deep learning methods for semantic segmentation in remote sensing with small data: A survey. Remote Sensing 15(20): 4987.

Yu, H., Hou, Y., Wang, F., Wang, J., Zhu, J. & Guo, J. 2024. MSSFNet: A multiscale spatial-spectral fusion network for extracting offshore floating raft aquaculture areas in multispectral remote sensing images. Sensors 24(16): 5220.

Zhang, W. & Wang, A. 2023. Research on semantic segmentation method of remote sensing image based on self-supervised learning. International Journal of Advanced Computer Science and Applications 14(8). https://doi.org/10.14569/IJACSA.2023.0140855.

Zheng, Y., Chen, Z., Zheng, T., Tian, C. & Dong, W. 2025. PSNet: A universal algorithm for multispectral remote sensing image segmentation. Remote Sensing 17(4): 563.

Zhu, H., Tan, R., Han, L., Fan, H., Wang, Z., Du, B., Liu, S. & Liu, Q. 2022. DSSM: A deep neural network with spectrum separable module for multi-spectral remote sensing image segmentation. Remote Sensing 14(4): 818.

 

*Corresponding author; email: raihanimohamed@upm.edu.my