Deep Learning (CNN) for Detecting Road Infrastructure in Old Mosul City Using High-Resolution Aerial Imagery

Authors

  • Mustafa Ismat Abdulrahman Department of Surveying Engineering Techniques, Technical College of Kirkuk, Northern Technical University Kirkuk, Iraq. https://orcid.org/0009-0008-7689-502X
  • Muntadher Aidi Shareef Department of Surveying Engineering Techniques, Technical College of Kirkuk, Northern Technical University Kirkuk, Iraq.
  • Alyaa Abbas Al-Attar Northern Technical University, Mosul, Iraq.

DOI:

https://doi.org/10.21123/bsj.2024.9449

Keywords:

CNN, Deep learning, Road detection, Road infrastructure, Old Mosul City

Abstract

Road networks and transportation infrastructure play a crucial role in many applications such as urban planning and environmental assessment. Remote sensing provides multispectral data that can be used to identify, extract, and map roads. However, accurate mapping of road networks from aerial imagery poses challenges due to the complexity of real-world road patterns. The aim of this study was to develop deep learning techniques for automated road extraction from aerial photographs. The study evaluated several convolutional neural network (CNN) architectures were including a hybrid spectral-spatial network (HybridSN). Models were assessed on a dataset of urban aerial images with lidar-derived ground truth labels. The joint modeling of multi-modal cues enables highly precise localization and delineation of road segments. The results of the HybridSN integrating both spectral and spatial processing achieved the top performance with 96.9% overall accuracy and 80.6% intersection-over-union after post-processing. In comparison, CNNs leveraging spatial context alone perform worse with the best overall accuracy of 95.4% after post-processing. The findings demonstrate the importance of fusing spectral and spatial data within deep learning frameworks for road extraction.

References

Jassim OA, Abed MJ, Saied ZH. Indoor/Outdoor Deep Learning Based Image Classification for Object Recognition Applications. Baghdad Sic J. 2023 Dec 5; 20(6 (Suppl.)): 2540. https://orcid.org/0009-0004-1607-2778

Abd Alsammed SM. Advanced GIS-based multi-function support system for identifying the best route. Baghdad Sic J. 2022 Jun 1; 19(3): 0631. https://doi.org/10.21123/bsj.2022.19.3.0631

Kadhim MA, Abed MH. Convolutional Neural Network for Satellite Image Classification. Intelligent Information and Database Systems: Recent Developments. 2020; 11: 165-178. https://doi.org/10.1007/978-3-030-14132-5_13

Zeng Y, Guo Y, Li J. Recognition and extraction of high-resolution satellite remote sensing image buildings based on deep learning. Neural Comput Appl. Feb 2022; 34(4): 2691-2706. https://doi.org/10.1007/s00521-021-06027-1

Shareef MA, Toumi A, Khenchaf A. Estimation of water quality parameters using the regression model with fuzzy k-means clustering. Int J Adv Comput Sci Appl. 2014; 5(6) : 151-157.

https://doi.org/10.14569/IJACSA.2014.050624

Keshk H, Yin X. Classification of EgyptSat-1 Images Using Deep Learning Methods. Int J Sens Wirel Commun Control. Feb 2020; 10: 37-46. https://doi.org/10.2174/2210327909666190207153858

Yu Y, Gong Z, Zhong P. An Unsupervised Convolutional Feature Fusion Network for Deep Representation of Remote Sensing Images. IEEE Trans Geosci Remote Sens. Dec 2017; 15: 23-27. https://doi.org/10.1109/LGRS.2017.2767626

Shareef MA, Ameen MH, Ajaj QM. Change detection and GIS-based fuzzy AHP to evaluate the degradation and reclamation land of Tikrit City Iraq. Geodesy Cartogr. Dec 2020; 46(4): 194-203. https://doi.org/10.3846/gac.2020.11616

Dai J, Du Y, Zhu T, Wang Y, Gao L. Multiscale Residual Convolution Neural Network and Sector Descriptor-Based Road Detection Method. IEEE Access. Nov 2019; 7: 173377-173392. https://doi.org/10.1109/ACCESS.2019.2956725

Wei Y, Zhang K, Ji S. Simultaneous Road Surface and Centerline Extraction From Large-Scale Remote Sensing Images Using CNN-Based Segmentation and Tracing. IEEE Trans Geosci Remote Sens. May 2020; 58(12): 8919-8931. https://doi.org/10.1109/TGRS.2020.2991733

Lu X, Zhong Y, Zheng Z, Liu Y, Zhao J, Ma A, et al. Multi-Scale and Multi-Task Deep Learning Framework for Automatic Road Extraction. IEEE Trans Geosci Remote Sens. Aug 2019; 57(11): 9362-9377. https://doi.org/10.1109/TGRS.2019.2926397

Han X, Lu J, Zhao C, You S, Li H. Semisupervised and akly Supervised Road Detection Based on Generative Adversarial Networks. IEEE Signal Process. Lett. Feb. 2018; 25(4): 551-555. https://doi.org/10.1109/LSP.2018.2809685

Li X, Wang Y, Zhang L, Liu S, Mei J, Li Y. Topology-Enhanced Urban Road Extraction via a Geographic Feature-Enhanced Network. IEEE Trans Geosci Remote Sens.May 2020; 58(12): 8819-8830. https://doi.org/10.1109/TGRS.2020.2991006

Sofla R, Alipour-Fard T, Arefi H. Road extraction from satellite and aerial image using SE-Unet. J Appl Remote Sens. . Jan 2021; 15(1): 014512 – 014512. https://doi.org/10.1117/1.JRS.15.014512

Fan R, Bocus MJ, Zhu Y, Jiao J, Wang L, Ma F, Cheng S, Liu M. Road crack detection using deep convolutional neural network and adaptive thresholding. In2019 IEEE Intelligent Vehicles Symposium (IV) 2019 Jun 9 (pp. 474-479). https://doi.org/10.1109/IVS.2019.8814000

Shi Q, Liu X, Li X. Road detection from remote sensing images by generative adversarial networks. IEEE access. 2017 Nov 13; 6: 25486-94.https://doi.org/10.1109/ACCESS.2017.2773142

Downloads

Issue

Section

article

How to Cite

1.
Deep Learning (CNN) for Detecting Road Infrastructure in Old Mosul City Using High-Resolution Aerial Imagery. Baghdad Sci.J [Internet]. [cited 2024 Nov. 21];22(5). Available from: https://bsj.uobaghdad.edu.iq/index.php/BSJ/article/view/9449