Abbas Mohammed Noori
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Abbas Mohammed Noori has a total of 26 co-authors. Here are the top co-authors by number of shared publications:
- Abdul Razzak Tarish Ziboon
Al-Esraa University College, Baghdad, Iraq
3 shared publications - Amjed Naser Al-Hameedawi
Civil Engineering Department, University of Technology, Baghdad 10066, Iraq
3 shared publications - Qayssar Mahmood Ajaj
Department of Surveying Engineering Technology, Technical Engineering College of Kirkuk, Northern Technical University, Kirkuk, Iraq
3 shared publications - Biswajeet Pradhan
Centre for Advanced Modelling and Geospatial Information Systems (CAMGIS), School of Civil, Environmental and Sociotechnical Engineering, University of Technology Sydney
2 shared publications Farman Galeb Saed
Technical College of Kirkuk, Northern Technical University
2 shared publications- Muntadher Aidi Shareef
Northern Technical University
2 shared publications
Latest Publications
- Open Access
Land-Cover Classification of Sentinel-2 Imagery Using a Hybrid Swin-Style Transformer and SegFormer-Inspired Network: A Case Study of Hawija, Iraq
Mapping land cover helps with the management of environmental, agricultural and urban areas; however, it is challenging in semi-arid regions, where several classes have similar spectral responses. The research offers a combination of new SWIN with SegFormer network for classifying five land-cover classes occurring in Hawija District, Iraq, using ten Sentinel-2 Level-2A spectral bands and eight spectral indices. As many as 111,740 labeled samples were extracted as 11 × 11-pixel patches with 18 input channels, and were divided by an 80:20 stratified random split into 80,452 training and 22,348 test samples. The data was divided into 80,452 training samples (72%), 8940 validation samples (8%), and 22,348 test samples (20%). The validation subset was used to monitor the training, early stopping, and checkpoint selection, while the test subset was only used for a performance evaluation at the end. The three-way partitioning protocol was adopted for the HSSNet Swin-style Transformer and SegFormer-like baseline. HSSNet is a combination of Swin-inspired attention with SegFormer-inspired feature processing, gated cross-attention fusion and squeeze-and-excitation channel recalibration. HSSNet obtained an overall accuracy (OA) of 96.18%, an average accuracy (AA) of 95.47% and a kappa coefficient of 0.9434, compared with 90.80%, 83.39% and 0.8614 for the Swin Transformer baseline and 93.68%, 90.73% and 0.9062 for the SegFormer-like baseline. The results indicate that hybrid feature fusion can improve land-cover discrimination in spectrally heterogeneous semi-arid landscapes, although spatially independent validation is still required to confirm the generalization of these findings.
Open AccessLand-Cover Classification of Sentinel-2 Imagery Using a Hybrid Swin-Style Transformer and SegFormer-Inspired Network: A Case Study of Hawija, Iraq
Mapping land cover helps with the management of environmental, agricultural and urban areas; however, it is challenging in semi-arid regions, where several classes have similar spectral responses. The research offers a combination of new SWIN with SegFormer network for classifying five land-cover classes occurring in Hawija District, Iraq, using ten Sentinel-2 Level-2A spectral bands and eight spectral indices. As many as 111,740 labeled samples were extracted as 11 × 11-pixel patches with 18 input channels, and were divided by an 80:20 stratified random split into 80,452 training and 22,348 test samples. The data was divided into 80,452 training samples (72%), 8940 validation samples (8%), and 22,348 test samples (20%). The validation subset was used to monitor the training, early stopping, and checkpoint selection, while the test subset was only used for a performance evaluation at the end. The three-way partitioning protocol was adopted for the HSSNet Swin-style Transformer and SegFormer-like baseline. HSSNet is a combination of Swin-inspired attention with SegFormer-inspired feature processing, gated cross-attention fusion and squeeze-and-excitation channel recalibration. HSSNet obtained an overall accuracy (OA) of 96.18%, an average accuracy (AA) of 95.47% and a kappa coefficient of 0.9434, compared with 90.80%, 83.39% and 0.8614 for the Swin Transformer baseline and 93.68%, 90.73% and 0.9062 for the SegFormer-like baseline. The results indicate that hybrid feature fusion can improve land-cover discrimination in spectrally heterogeneous semi-arid landscapes, although spatially independent validation is still required to confirm the generalization of these findings.
- Open Access
- 10 July 2025
- Published by MDPI AG in Applied Sciences
- Vol. 15 (14) , 7770
Deep-Learning Integration of CNN–Transformer and U-Net for Bi-Temporal SAR Flash-Flood Detection
Flash floods are natural disasters that have significant impacts on human life and economic damage. The detection of flash floods using remote-sensing techniques provides essential data for subsequent flood-risk assessment through the preparation of flood inventory samples. In this research, a new deep-learning approach for bi-temporal flash-flood detection in Synthetic Aperture Radar (SAR) is proposed. It combines a U-Net convolutional network with a Transformer model using a compact Convolutional Tokenizer (CCT) to improve the efficiency of long-range dependency learning. The hybrid model, namely CCT-U-ViT, naturally combines the spatial feature extraction of U-Net and the global context capability of Transformer. The model significantly reduces the number of basic blocks as it uses the CCT tokenizer instead of conventional Vision Transformer tokenization, which makes it the right fit for small flood detection datasets. This model improves flood boundary delineation by involving local spatial patterns and global contextual relations. However, the method is based on Sentinel-1 SAR images and focuses on Erbil, Iraq, which experienced an extreme flash flood in December 2021. The experimental comparison results show that the proposed CCT-U-ViT outperforms multiple baseline models, such as conventional CNNs, U-Net, and Vision Transformer, obtaining an impressive overall accuracy of 91.24%. Furthermore, the model obtains better precision and recall with an F1-score of 91.21% and mIoU of 83.83%. Qualitative results demonstrate that CCT-U-ViT can effectively preserve the flood boundaries with higher precision and less salt-and-pepper noise compared with the state-of-the-art approaches. This study underscores the significance of hybrid deep-learning models in enhancing the precision of flood detection with SAR data, providing valuable insights for the advancement of real-time flood monitoring and risk management systems.
- 10 July 2025
- Published by MDPI AG in Applied Sciences
- Vol. 15 (14) , 7770
Open AccessDeep-Learning Integration of CNN–Transformer and U-Net for Bi-Temporal SAR Flash-Flood Detection
Flash floods are natural disasters that have significant impacts on human life and economic damage. The detection of flash floods using remote-sensing techniques provides essential data for subsequent flood-risk assessment through the preparation of flood inventory samples. In this research, a new deep-learning approach for bi-temporal flash-flood detection in Synthetic Aperture Radar (SAR) is proposed. It combines a U-Net convolutional network with a Transformer model using a compact Convolutional Tokenizer (CCT) to improve the efficiency of long-range dependency learning. The hybrid model, namely CCT-U-ViT, naturally combines the spatial feature extraction of U-Net and the global context capability of Transformer. The model significantly reduces the number of basic blocks as it uses the CCT tokenizer instead of conventional Vision Transformer tokenization, which makes it the right fit for small flood detection datasets. This model improves flood boundary delineation by involving local spatial patterns and global contextual relations. However, the method is based on Sentinel-1 SAR images and focuses on Erbil, Iraq, which experienced an extreme flash flood in December 2021. The experimental comparison results show that the proposed CCT-U-ViT outperforms multiple baseline models, such as conventional CNNs, U-Net, and Vision Transformer, obtaining an impressive overall accuracy of 91.24%. Furthermore, the model obtains better precision and recall with an F1-score of 91.21% and mIoU of 83.83%. Qualitative results demonstrate that CCT-U-ViT can effectively preserve the flood boundaries with higher precision and less salt-and-pepper noise compared with the state-of-the-art approaches. This study underscores the significance of hybrid deep-learning models in enhancing the precision of flood detection with SAR data, providing valuable insights for the advancement of real-time flood monitoring and risk management systems.
- Open Access
- 17 June 2025
- Published by MDPI AG in Preprints.org
Deep Learning Integration of CNN-Transformer and UNet for Bi-Temporal SAR Flash Flood Detection
A new deep learning approach for bi-temporal flash flood detection in synthetic aperture radar (SAR) is proposed in this research. It combines a U-Net convolutional network with a Transformer model using a compact Convolutional Tokenizer (CCT) to improve the efficiency of long-range dependency learning. The hybrid model, namely CCT-U-ViT, that naturally combines the spatial feature extraction of U-Net and the global context capability of Transformer. The model significantly reduces the number of basic blocks as it uses the CCT tokenizer instead of conventional Vision Transformer tokenization, which makes it the right fit for small flood detection datasets. This model improves flood boundary delineation by involving local spatial patterns and global contextual relations. However, the method is based on Sentinel-1 SAR images and focuses on Erbil, Iraq, that experienced an extreme flash flood in December 2021. The experimental comparison results show that the proposed CCT-U-ViT outperforms multiple baseline models, such as conventional CNNs, U-Net, and Vision Transformer, obtaining an impressive overall accuracy of 91.24%. In addition, Furthermore, the model gets better precision and recall with F1-score of 91.21% and mIoU of 83.83%. Qualitative results demonstrate that CCT-U-ViT can effectively preserve the flood boundaries with higher precision and less salt-and-pepper noise compared with the state-of-the-art approaches. This study underscores the significance of hybrid deep learning models in enhancing the precision of flood detection with SAR data, providing valuable insights for the advancement of real-time flood monitoring and crisis management systems.
- 17 June 2025
- Published by MDPI AG in Preprints.org
Open AccessDeep Learning Integration of CNN-Transformer and UNet for Bi-Temporal SAR Flash Flood Detection
A new deep learning approach for bi-temporal flash flood detection in synthetic aperture radar (SAR) is proposed in this research. It combines a U-Net convolutional network with a Transformer model using a compact Convolutional Tokenizer (CCT) to improve the efficiency of long-range dependency learning. The hybrid model, namely CCT-U-ViT, that naturally combines the spatial feature extraction of U-Net and the global context capability of Transformer. The model significantly reduces the number of basic blocks as it uses the CCT tokenizer instead of conventional Vision Transformer tokenization, which makes it the right fit for small flood detection datasets. This model improves flood boundary delineation by involving local spatial patterns and global contextual relations. However, the method is based on Sentinel-1 SAR images and focuses on Erbil, Iraq, that experienced an extreme flash flood in December 2021. The experimental comparison results show that the proposed CCT-U-ViT outperforms multiple baseline models, such as conventional CNNs, U-Net, and Vision Transformer, obtaining an impressive overall accuracy of 91.24%. In addition, Furthermore, the model gets better precision and recall with F1-score of 91.21% and mIoU of 83.83%. Qualitative results demonstrate that CCT-U-ViT can effectively preserve the flood boundaries with higher precision and less salt-and-pepper noise compared with the state-of-the-art approaches. This study underscores the significance of hybrid deep learning models in enhancing the precision of flood detection with SAR data, providing valuable insights for the advancement of real-time flood monitoring and crisis management systems.
- Open Access
- 23 September 2024
- Published by University of Technology, Iraq - DIGITAL COMMONS in Engineering and Technology Journal
- p. 1-9
Assessment of flash flood detection in Erbil city using change detection indices for SAR images
The frequency and intensity of flash floods are expected to increase due to climate change, resulting in significant casualties and damage to infrastructure and the economy. Reliable and timely flood maps are essential for an effective disaster management plan. On December 17, 2021, a severe flash flood in Erbil City resulted in twelve fatalities and extensive damage to the affected area. The ability of synthetic aperture radar (SAR) images to penetrate clouds and heavy precipitation is vital for accurate flood imaging. This study analyzed Sentinel-1 satellite-based radar images before and after the flood to detect the inundation. Two change detection techniques, the Normalized Change Index (NCI) and the Ratio Index (RI), combined with semi-automatic thresholding were employed. Both methods successfully identified the flooded region with consistent findings. The overall accuracy of NCI and RI were 90.5% and 84.3% respectively. The accuracy assessment revealed that the NCI method outperforms the RI method in detecting the flash flood event in Erbil City. This model is viable for disaster management, enabling the evaluation of damage to critical municipal infrastructure and other assets, thus supporting effective urban governance and timely response to emergencies. At the final stage of disaster management, this model can be implemented to evaluate the extent of loss on major municipal infrastructure and other properties within the area.
- 23 September 2024
- Published by University of Technology, Iraq - DIGITAL COMMONS in Engineering and Technology Journal
- p. 1-9
Open AccessAssessment of flash flood detection in Erbil city using change detection indices for SAR images
The frequency and intensity of flash floods are expected to increase due to climate change, resulting in significant casualties and damage to infrastructure and the economy. Reliable and timely flood maps are essential for an effective disaster management plan. On December 17, 2021, a severe flash flood in Erbil City resulted in twelve fatalities and extensive damage to the affected area. The ability of synthetic aperture radar (SAR) images to penetrate clouds and heavy precipitation is vital for accurate flood imaging. This study analyzed Sentinel-1 satellite-based radar images before and after the flood to detect the inundation. Two change detection techniques, the Normalized Change Index (NCI) and the Ratio Index (RI), combined with semi-automatic thresholding were employed. Both methods successfully identified the flooded region with consistent findings. The overall accuracy of NCI and RI were 90.5% and 84.3% respectively. The accuracy assessment revealed that the NCI method outperforms the RI method in detecting the flash flood event in Erbil City. This model is viable for disaster management, enabling the evaluation of damage to critical municipal infrastructure and other assets, thus supporting effective urban governance and timely response to emergencies. At the final stage of disaster management, this model can be implemented to evaluate the extent of loss on major municipal infrastructure and other properties within the area.
- 1 January 2024
- Published by AIP Publishing in AIP Conference Proceedings
- Vol. 3249 (1) , 050005
Monitoring and prediction of land cover changes of Kirkuk City using machine learning and remote sensing data
Expansion of land cover (LC) is significantly impacted by man-made activities, particularly in areas that are expanding quickly. A CA-ANN model, which stands fo Expansion of land cover (LC) is significantly impacted by man-made activities, particularly in areas that are expanding quickly. A CA-ANN model, which stands fo
- 1 January 2024
- Published by AIP Publishing in AIP Conference Proceedings
- Vol. 3249 (1) , 050005
Monitoring and prediction of land cover changes of Kirkuk City using machine learning and remote sensing data
Expansion of land cover (LC) is significantly impacted by man-made activities, particularly in areas that are expanding quickly. A CA-ANN model, which stands fo Expansion of land cover (LC) is significantly impacted by man-made activities, particularly in areas that are expanding quickly. A CA-ANN model, which stands fo
- 1 January 2024
- Published by AIP Publishing in AIP Conference Proceedings
- Vol. 3249 (1) , 050004
Flash flood susceptibility mapping via morphometric analysis of Erbil City basins, Iraq
The characteristics of a watershed and how it responds to hydrological processes can significantly impact the occurrence and extent of floods. In the last decad The characteristics of a watershed and how it responds to hydrological processes can significantly impact the occurrence and extent of floods. In the last decad
- 1 January 2024
- Published by AIP Publishing in AIP Conference Proceedings
- Vol. 3249 (1) , 050004
Flash flood susceptibility mapping via morphometric analysis of Erbil City basins, Iraq
The characteristics of a watershed and how it responds to hydrological processes can significantly impact the occurrence and extent of floods. In the last decad The characteristics of a watershed and how it responds to hydrological processes can significantly impact the occurrence and extent of floods. In the last decad
- Open Access
- 7 June 2022
- Published by Wiley in Journal of Sensors
- Vol. 2022, 1-11
Earthquake-Induced Ground Deformation Assessment via Sentinel-1 Radar Aided at Darbandikhan Town
Deformation monitoring has conventional increasing consideration in recent years due to its great importance in modern engineering. Darbandikhan town was hit by a strong earthquake on 12th November 2017 which is the Sarpol-e Zahab earthquake. Interferometric Synthetic Aperture Radar (InSAR) technique has been used to investigate Darbandikhan town stability and the surrounding areas of the Kurdistan region in Iraq. The purpose of this research is to estimate the ground vertical displacement induced by the earthquake as well as the east-west deformation components. Moreover, previous research covered more of Irans side country that the earthquake had an impact on with little on Iraqs side (Northern-eastern of Iraq), so it was worth researching on Darbandikhan town and surrounding structures to estimate the displacement on its infrastructures. For this study, two pairs of Sentinel-1 Synthetic Aperture Radar (SAR) images are used from each ascending and descending tracks over the study area to validate the work. Then, two differential interferograms are created from these images, along Darbandikhan town and the surrounding areas. These differential interferograms are then converted to line of sight (LOS) displacements using open source software (SNAP 8.0). The overall vertical displacement of the residential area of Darbandikhan was estimated from -4 to -10cm after the earthquake, and the overall east-west displacement was ranging from 2 to 6.4cm. The current study does not only examine the northern east of Iraq but the whole area of Iran and Iraq in general that lies within the influence of the earthquake. This research could be useful to estimate the overall deformation of Darbandikhan town in particular and the whole area in general. Therefore, the authorities and civil engineers could think of more sustainable buildings and structures before they start to plan within the area of earthquake impact.
- 7 June 2022
- Published by Wiley in Journal of Sensors
- Vol. 2022, 1-11
Open AccessEarthquake-Induced Ground Deformation Assessment via Sentinel-1 Radar Aided at Darbandikhan Town
Deformation monitoring has conventional increasing consideration in recent years due to its great importance in modern engineering. Darbandikhan town was hit by a strong earthquake on 12th November 2017 which is the Sarpol-e Zahab earthquake. Interferometric Synthetic Aperture Radar (InSAR) technique has been used to investigate Darbandikhan town stability and the surrounding areas of the Kurdistan region in Iraq. The purpose of this research is to estimate the ground vertical displacement induced by the earthquake as well as the east-west deformation components. Moreover, previous research covered more of Irans side country that the earthquake had an impact on with little on Iraqs side (Northern-eastern of Iraq), so it was worth researching on Darbandikhan town and surrounding structures to estimate the displacement on its infrastructures. For this study, two pairs of Sentinel-1 Synthetic Aperture Radar (SAR) images are used from each ascending and descending tracks over the study area to validate the work. Then, two differential interferograms are created from these images, along Darbandikhan town and the surrounding areas. These differential interferograms are then converted to line of sight (LOS) displacements using open source software (SNAP 8.0). The overall vertical displacement of the residential area of Darbandikhan was estimated from -4 to -10cm after the earthquake, and the overall east-west displacement was ranging from 2 to 6.4cm. The current study does not only examine the northern east of Iraq but the whole area of Iran and Iraq in general that lies within the influence of the earthquake. This research could be useful to estimate the overall deformation of Darbandikhan town in particular and the whole area in general. Therefore, the authorities and civil engineers could think of more sustainable buildings and structures before they start to plan within the area of earthquake impact.
- 1 July 2019
- Published by Elsevier in Journal of Hydrology
- Vol. 574, 964-979
Dam site suitability assessment at the Greater Zab River in northern Iraq using remote sensing data and GIS
- 1 July 2019
- Published by Elsevier in Journal of Hydrology
- Vol. 574, 964-979
Dam site suitability assessment at the Greater Zab River in northern Iraq using remote sensing data and GIS