Skip to main content

Advertisement

SpringerLink
Log in

Modeling PolSAR classification using convolutional neural network with homogeneity based kernel selection

  • Original Article
  • Published:
Modeling Earth Systems and Environment Aims and scope Submit manuscript

Abstract

Synthetic Aperture Radar (SAR) data interpretation is a well-covered study in remote sensing for various applications. The scattering information obtained by SAR data varies depending on the target location and texture. For instance, scattering information from an urban region is more heterogeneous than scattering information from vegetation or water. The homogeneity feature of the Gray Level Co-occurrence Matrix (GLCM) plays a vital role in determining the land cover regions. The modelling of land cover classification incorporates the most famous Convolutional Neural Network (CNN) model. Modeling classification using CNN requires the selection of the appropriate kernel size as it is an influential parameter for improving classification accuracy. Up until now, researchers worked on a single kernel size, while in this research, a multi-size kernel is used based on the homogeneity of an image patch. A pair of CNN models, each with a different kernel size, is being used. One CNN model classifies an image patch with high homogeneity, and another model classifies an image patch with low homogeneity. Experimental results show that the use of the two CNN models with varied kernels improves the accuracy of classification for both compact and full polarimetric data sets.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10
Fig. 11
Fig. 12
Fig. 13
Fig. 14

Similar content being viewed by others

Data availability

RISAT-1 dataset was received from SAC-ISRO, Ahmedabad and Flevoland dataset was available at the link provided in the article.

References

  • Biswas S, Sinha M (2021) Performances of deep learning models for indian ocean wind speed prediction. Modeling Earth Systems and Environment 7(2):809–831

    Article  Google Scholar 

  • Casallas A, Ferro C, Celis N, et al (2021) Long short-term memory artificial neural network approach to forecast meteorology and pm2. 5 local variables in bogotá, colombia. Modeling Earth Systems and Environment pp 1–14

  • Chen S, Jin M, Ding J (2021) Hyperspectral remote sensing image classification based on dense residual three-dimensional convolutional neural network. Multimedia Tools and Applications 80(2):1859–1882

    Article  Google Scholar 

  • Cheng J, Zhang F, Xiang D et al (2021) Polsar image land cover classification based on hierarchical capsule network. Remote Sensing 13(16):3132

    Article  Google Scholar 

  • Chollet F, et al (2018) Keras: The python deep learning library. Astrophysics source code library pp ascl–1806

  • Dasari K, Lokam A (2018) Exploring the capability of compact polarimetry (hybrid pol) c band risat-1 data for land cover classification. IEEE Access 6:57,981–57,993

  • De S, Bruzzone L, Bhattacharya A et al (2017) A novel technique based on deep learning and a synthetic target database for classification of urban areas in polsar data. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 11(1):154–170

    Article  Google Scholar 

  • Deng X, López-Martínez C, Chen J et al (2017) Statistical modeling of polarimetric sar data: A survey and challenges. Remote Sensing 9(4):348

    Article  Google Scholar 

  • ESA (2018) Esa earth observation data. Retrieved from https://earth.esa.int/web/polsarpro/datasources/sample-datasets

  • Gao F, Huang T, Wang J et al (2017) Dual-branch deep convolution neural network for polarimetric sar image classification. Applied Sciences 7(5):447

    Article  Google Scholar 

  • Haralick RM, Shanmugam K, Dinstein IH (1973) Textural features for image classification. IEEE Transactions on systems, man, and cybernetics 6:610–621

    Article  Google Scholar 

  • Jiang H, Lu N (2018) Multi-scale residual convolutional neural network for haze removal of remote sensing images. Remote Sensing 10(6):945

    Article  Google Scholar 

  • Johansson JO (2000) Measuring homogeneity of planar point-patterns by using kurtosis. Pattern Recognition Letters 21(13–14):1149–1156

    Article  Google Scholar 

  • Khanday AMUD, Bhushan B, Jhaveri RH, et al (2022) Nnpcov19: Artificial neural network-based propaganda identification on social media in covid-19 era. Mobile Information Systems 2022

  • Knauer U, Meffert B (2010) Fast computation of region homogeneity with application in a surveillance task. In: Proceedings of ISPRS Commission V Mid-Term Symposium Close Range Image Measurement Techniques, pp 337–342

  • Lee JS, Pottier E (2017) Polarimetric radar imaging: from basics to applications. CRC Press

    Book  Google Scholar 

  • Lei R, Zhang C, Du S et al (2021) A non-local capsule neural network for hyperspectral remote sensing image classification. Remote Sensing Letters 12(1):40–49

    Article  Google Scholar 

  • Li M, Li X, Sun W et al (2021) Efficient convolutional neural network with multi-kernel enhancement features for real-time facial expression recognition. Journal of Real-Time Image Processing 18(6):2111–2122

    Article  Google Scholar 

  • Liu X, Jiao L, Tang X et al (2018) Polarimetric convolutional network for polsar image classification. IEEE Transactions on Geoscience and Remote Sensing 57(5):3040–3054

    Article  Google Scholar 

  • Ma Y, Li Y, Zhu L (2019) Land cover classification for polarimetric sar image using convolutional neural network and superpixel. Progress In Electromagnetics Research B 83:111–128

    Article  Google Scholar 

  • Masjedi A, Zoej MJV, Maghsoudi Y (2015) Classification of polarimetric sar images based on modeling contextual information and using texture features. IEEE Transactions on Geoscience and Remote Sensing 54(2):932–943

    Article  Google Scholar 

  • Momeneh S, Nourani V (2022) Application of a novel technique of the multi-discrete wavelet transforms in hybrid with artificial neural network to forecast the daily and monthly streamflow. Modeling Earth Systems and Environment pp 1–20

  • Namburu A, Sumathi D, Raut R, et al (2022) Fpga-based deep learning models for analysing corona using chest x-ray images. Mobile Information Systems 2022

  • Parikh HI, Patel SB, Patel VD (2019) Land cover classification using ensemble techniques. In: International Conference on Pattern Recognition and Machine Intelligence, Springer, pp 340–349

  • Pottier E, Ferro-Famil L (2012) Polsarpro v5.0: An esa educational toolbox used for self-education in the field of polsar and pol-insar data analysis. pp 7377–7380, 10.1109/IGARSS.2012.6351925

  • Pulido D (2020) Automated extraction of homogeneous regions by seeded region shrinkage. Journal of Applied Remote Sensing 14(3):036,518

  • Raney RK (2007) Hybrid-polarity sar architecture. IEEE Transactions on Geoscience and Remote Sensing 45(11):3397–3404

    Article  Google Scholar 

  • Raney RK, Cahill JT, Patterson GW, et al (2012) The m-chi decomposition of hybrid dual-polarimetric radar data with application to lunar craters. Journal of Geophysical Research: Planets 117(E12)

  • Sasaki Y (2007) The truth of the f-measure. Teach Tutor Mater

  • Shakya A, Biswas M, Pal M (2021) Parametric study of convolutional neural network based remote sensing image classification. International Journal of Remote Sensing 42(7):2663–2685

    Article  Google Scholar 

  • Sprunger C, Muther T, Syed FI et al (2022) State of the art progress in hydraulic fracture modeling using ai/ml techniques. Modeling Earth Systems and Environment 8(1):1–13

    Article  Google Scholar 

  • Srinivasu PN, Sandhya N, Jhaveri RH, et al (2022) From blackbox to explainable ai in healthcare: Existing tools and case studies. Mobile Information Systems 2022

  • Zayed N, Elnemr HA (2015) Statistical analysis of haralick texture features to discriminate lung abnormalities. International journal of biomedical imaging 2015

  • Zhai W, Shen H, Huang C et al (2016) Fusion of polarimetric and texture information for urban building extraction from fully polarimetric sar imagery. Remote Sensing Letters 7(1):31–40

    Article  Google Scholar 

  • Zhang Z, Wang H, Xu F et al (2017) Complex-valued convolutional neural network and its application in polarimetric sar image classification. IEEE Transactions on Geoscience and Remote Sensing 55(12):7177–7188

    Article  Google Scholar 

Download references

Acknowledgements

Authors are thankful to the SAC-ISRO, Ahmedabad for providing RISAT-1 dataset and to NASA/JPL for making the Flevoland dataset available for research. Authors are also thankful to the developers of PolSARPro and SNAP software. Authors are also thankful to NRSC, India for making available Bhuvan thematic map.

Author information

Authors and Affiliations

  1. Computer Engineering Department, School of Technology, Pandit Deendayal Energy University, Gandhinagar, Gujarat, India

    Hemani Parikh & Samir Patel

  2. Information Technology Department, Vishwakarma Government Engineering College, Chandkheda, Ahmedabad, Gujarat, India

    Vibha Patel

Authors
  1. Hemani Parikh
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Samir Patel
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Vibha Patel
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Hemani Parikh.

Ethics declarations

Conflict of interest

On behalf of all authors, the corresponding author states that there is no conflict of interest.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Parikh, H., Patel, S. & Patel, V. Modeling PolSAR classification using convolutional neural network with homogeneity based kernel selection. Model. Earth Syst. Environ. 9, 3801–3813 (2023). https://doi.org/10.1007/s40808-023-01700-x

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s40808-023-01700-x

Keywords

Search

Navigation