Sparsity enhanced MRF algorithm for automatic object detection in GPR imagery

Changpu Meng; Jie Yang; Changpu Meng; Jie Yang

doi:10.3934/mbe.2023707

Mathematical Biosciences and Engineering

2023, Volume 20, Issue 9: 15883-15897. doi: 10.3934/mbe.2023707

Previous Article Next Article

Research article Special Issues

Sparsity enhanced MRF algorithm for automatic object detection in GPR imagery

Changpu Meng ^{1
,
,},
Jie Yang ²

1.
Corporation of iFLYTEK Co., Ltd., Hefei, 230088, R.P. China
2.
School of Computing and Information Technology, Faculty of Engineering and Information Sciences, University of Wollongong, Northfields Avenue, Wollongong, 2522, Australia

Received: 11 April 2023 Revised: 30 June 2023 Accepted: 13 July 2023 Published: 01 August 2023

This study addressed the problem of automated object detection from ground penetrating radar imaging (GPR), using the concept of sparse representation. The detection task is first formulated as a Markov random field (MRF) process. Then, we propose a novel detection algorithm by introducing the sparsity constraint to the standard MRF model. Specifically, the traditional approach finds it difficult to determine the central target due to the influence of different neighbors from the imaging area. As such, we introduce a domain search algorithm to overcome this issue and increase the accuracy of target detection. Additionally, in the standard MRF model, the Gibbs parameters are empirically predetermined and fixed during the detection process, yet those hyperparameters may have a significant effect on the performance of the detection. Accordingly, in this paper, Gibbs parameters are self-adaptive and fine-tuned using an iterative updating strategy followed the concept of sparse representation. Furthermore, the proposed algorithm has then been proven to have a strong convergence property theoretically. Finally, we verify the proposed method using a real-world dataset, with a set of ground penetrating radar antennas in three different transmitted frequencies (50 MHz, 200 MHz and 300 MHz). Experimental evaluations demonstrate the advantages of utilizing the proposed algorithm to detect objects in ground penetrating radar imagery, in comparison with four traditional detection algorithms.
- Ground penetrating radar (GPR) imaging,
- object detection,
- Markov random field,
- sparse representation
Citation: Changpu Meng, Jie Yang. Sparsity enhanced MRF algorithm for automatic object detection in GPR imagery[J]. Mathematical Biosciences and Engineering, 2023, 20(9): 15883-15897. doi: 10.3934/mbe.2023707

Related Papers:

Abstract

This study addressed the problem of automated object detection from ground penetrating radar imaging (GPR), using the concept of sparse representation. The detection task is first formulated as a Markov random field (MRF) process. Then, we propose a novel detection algorithm by introducing the sparsity constraint to the standard MRF model. Specifically, the traditional approach finds it difficult to determine the central target due to the influence of different neighbors from the imaging area. As such, we introduce a domain search algorithm to overcome this issue and increase the accuracy of target detection. Additionally, in the standard MRF model, the Gibbs parameters are empirically predetermined and fixed during the detection process, yet those hyperparameters may have a significant effect on the performance of the detection. Accordingly, in this paper, Gibbs parameters are self-adaptive and fine-tuned using an iterative updating strategy followed the concept of sparse representation. Furthermore, the proposed algorithm has then been proven to have a strong convergence property theoretically. Finally, we verify the proposed method using a real-world dataset, with a set of ground penetrating radar antennas in three different transmitted frequencies (50 MHz, 200 MHz and 300 MHz). Experimental evaluations demonstrate the advantages of utilizing the proposed algorithm to detect objects in ground penetrating radar imagery, in comparison with four traditional detection algorithms.

References

[1]	E. Pasolli, F. Melgani, M. Donelli, Automatic Analysis of GPR Images: A Pattern-Recognition Approach, IEEE Transact. Geosci. Remote Sens., 47 (2009), 2206–2217. https://doi.org/10.1109/TGRS.2009.2012701 doi: 10.1109/TGRS.2009.2012701
[2]	H. Harkat, A. E. Ruano, M. G. Ruano, S. D. Bennani, GPR target detection using a neural network classifier designed by a multi-objective genetic algorithm, Appl. Soft Comput., 79 (2019), 310–325. https://doi.org/10.1016/j.asoc.2019.03.030 doi: 10.1016/j.asoc.2019.03.030
[3]	U. Pe'er, J. G. Dy, Automated Target Detection for Geophysical Applications, IEEE Transact. Geosci. Remote Sens., 55 (2017), 1563–1572. https://doi.org/10.1109/TGRS.2016.2627245 doi: 10.1109/TGRS.2016.2627245
[4]	R. Sakaguchi, K. D. Morton, L. M. Collins, P. A. Torrione, A Comparison of Feature Representations for Explosive Threat Detection in Ground Penetrating Radar Data, IEEE Transact. Geosci. Remote Sens., 55 (2017), 6736–6745. https://doi.org/10.1109/TGRS.2017.2732226 doi: 10.1109/TGRS.2017.2732226
[5]	T. N. Tran, R. Wehrens, D. H. Hoekman, L. M. C. Buydens, Initialization of Markov random field clustering of large remote sensing images, IEEE Transact. Geosci. Remote Sens., 43 (2005), 1912–1919. https://doi.org/10.1109/TGRS.2005.848427 doi: 10.1109/TGRS.2005.848427
[6]	A. Bouzerdoum, J. Yang, F. Tivive, Compressive sensing for multipolarization through-the-wall radar imaging, Compressive Sensing for Urban Radar, Ed. M. G. Amin, United States: CRC Press, (2014), 231–250. http://dx.doi.org/10.1201/b17252-7
[7]	J. Yang, A. Bouzerdoum, S. L. Phung, A Neural Network pruning approach based on Compressive Sampling, in 2009 International Joint Conference on Neural Networks, (2009), 3428–3435. https://doi.org/10.1109/IJCNN.2009.5179045
[8]	J. Yang, J. Ma, M. J. Berryman, P. Perez, A structure optimization algorithm of neural networks for large-scale data sets, in 2014 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE), (2014), 956–961. https://doi.org/10.1109/FUZZ-IEEE.2014.6891662
[9]	H. Liu, Y. Yue, C. Liu, B. F. Spencer, J. Cui, Automatic recognition and localization of underground pipelines in GPR B-scans using a deep learning model, Tunnell. Underground Space Technol., 134 (2023), 104861. https://doi.org/10.1016/j.tust.2022.104861 doi: 10.1016/j.tust.2022.104861
[10]	S. Goodarzi, H. F. Kashani, A. Saeedi, J. Oke, C. L. Ho, Stochastic analysis for estimating track geometry degradation rates based on GPR and LiDAR data, Construct. Building Mater., 369 (2023), 130591. https://doi.org/10.1016/j.conbuildmat.2023.130591 doi: 10.1016/j.conbuildmat.2023.130591
[11]	M. Gaballah, T. Alharbi, 3-D GPR visualization technique integrated with electric resistivity tomography for characterizing near-surface fractures and cavities in limestone, J. Taibah Univer. Sci., 16 (2022), 224–239. https://doi.org/10.1080/16583655.2022.2040242 doi: 10.1080/16583655.2022.2040242
[12]	F. Bandini, L. Kooij, B. k. Mortensen, M. B. Caspersen, L. G. Thomsen, D. Olesen, P, et al., Mapping inland water bathymetry with Ground Penetrating Radar (GPR) on board Unmanned Aerial Systems (UASs), J. Hydrol., 616 (2023), 128789. https://doi.org/10.1016/j.jhydrol.2022.128789 doi: 10.1016/j.jhydrol.2022.128789
[13]	Y. Wu, K. Ji, W. Yu, Y. Su, Region-Based Classification of Polarimetric SAR Images Using Wishart MRF, IEEE Geosci. Remote Sens. Letters, 5 (2008), 668–672. https://doi.org/10.1109/LGRS.2008.2002263 doi: 10.1109/LGRS.2008.2002263
[14]	M. Gong, L. Su, M. Jia, W. Chen, Fuzzy Clustering With a Modified MRF Energy Function for Change Detection in Synthetic Aperture Radar Images, IEEE Transact. Fuzzy Syst., 22 (2014), 98–109. https://doi.org/10.1109/TFUZZ.2013.2249072 doi: 10.1109/TFUZZ.2013.2249072
[15]	Y. Yang, X. Cong, K. Long, Y. Luo, W. Xie, Qun Wan, MRF model-based joint interrupted SAR imaging and coherent change detection via variational Bayesian inference, Signal Process., 151 (2018), 144–154. https://doi.org/10.1016/j.sigpro.2018.05.007 doi: 10.1016/j.sigpro.2018.05.007
[16]	M. Liu, Y. Deng, C. Han, W. Hou, Y. Gao, C. Wang, et al., An Innovative Supervised Classification Algorithm for PolSAR Image Based on Mixture Model and MRF, Remote Sens., 14 (2022), 5506–5506. https://doi.org/10.3390/rs14215506 doi: 10.3390/rs14215506
[17]	F. Houcemeddine, K. Karim, Image segmentation using MRF model optimized by a hybrid ACO-ICM algorithm, Soft Comput., 25 (2021), 10181–10204. https://doi.org/10.1007/s00500-021-05957-1 doi: 10.1007/s00500-021-05957-1
[18]	D. L. Donoho, M. Elad, V. N. Temlyakov, Stable recovery of sparse overcomplete representations in the presence of noise, IEEE Transact. Inform. Theory, 52 (2006), 6–18. https://doi.org/10.1109/TIT.2005.860430 doi: 10.1109/TIT.2005.860430
[19]	M. Aharon, M. Elad, A. Bruckstein, K-SVD: An Algorithm for Designing Overcomplete Dictionaries for Sparse Representation, IEEE Transact. Signal Process., 54 (2006), 4311–4322. https://doi.org/10.1109/TSP.2006.881199 doi: 10.1109/TSP.2006.881199
[20]	S. Agarwal, D. Roth, Learning a sparse representation for object detection, 7$^{th}$ edition, European Conference on Computer Vision Copenhagen, ECCV 2002- Copenhagen, Denmark, 2002,113–127. https://doi.org/10.1007/3-540-47979-1_8
[21]	K. Huang, S. Aviyente, Sparse representation for signal classification, Adv. Neural Inform. Process. Syst., 19 (2006), 609–-616.
[22]	R. Rubinstein, A. M. Bruckstein, M. Elad, Dictionaries for Sparse Representation Modeling, Proceed. IEEE, 98 (2010), 1045–1057. https://doi.org/10.1109/JPROC.2010.2040551 doi: 10.1109/JPROC.2010.2040551
[23]	W. Dong, L. Zhang, G. Shi, X. Li, Nonlocally Centralized Sparse Representation for Image Restoration, IEEE Transact. Image Process., 22 (2013), 1620–1630. https://doi.org/10.1109/TIP.2012.2235847 doi: 10.1109/TIP.2012.2235847
[24]	A. Blake, P. Kohli, C. Rother, Markov Random Fields for Vision and Image Processing, The MIT Press, USA, 2011.
[25]	G. Andrew, J. Gao, Scalable Training of L1-Regularized Log-Linear Models, in International Conference on Machine Learning, 6 (2007), 33–40. https://doi.org/10.1145/1273496.1273501
[26]	L. Liu, Z. Jia, J. Yang, N. K. Kasabov, SAR Image Change Detection Based on Mathematical Morphology and the K-Means Clustering Algorithm, IEEE Access, 7 (2019), 43970–43978. https://doi.org/10.1109/ACCESS.2019.2908282 doi: 10.1109/ACCESS.2019.2908282
[27]	H. Zhu, W. Huang, H. Liu, Loess terrain segmentation from digital elevation models based on the region growth method, Phys. Geography, 39 (2018), 51–66. https://doi.org/10.1080/02723646.2017.1342215 doi: 10.1080/02723646.2017.1342215
[28]	M. Gong, Y. Liang, J. Shi, W. Ma, J. Ma, Fuzzy C-Means Clustering With Local Information and Kernel Metric for Image Segmentation, IEEE Transact. Image Process., 22 (2013), 573–584. https://doi.org/10.1109/TIP.2012.2219547 doi: 10.1109/TIP.2012.2219547

Reader Comments

Your name:*

Email:*
© 2023 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)