ASCEND-UNet: An Improved UNet Configuration Optimized for Rural Settlements Mapping
Abstract
:1. Introduction
2. Methods
2.1. Study Area
2.2. Data and Preprocessing
2.3. Model Design for ASCEND-UNet
2.4. HDC Block
2.5. scSE Block
2.6. ASPP Block
2.7. Model Training and Prediction
2.8. Accuracy Assessment
3. Results
3.1. Results and Comparison
3.2. Ablation Experiments
4. Discussions
4.1. Distinguishing Different Types of Rural Settlements
4.2. UNet Framework and Improvement
4.3. Multiple Dilated Convolution Block with Attention Mechanism
4.4. Limitations of the Proposed Model
5. Conclusions
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Acknowledgments
Conflicts of Interest
References
- Long, H.; Liu, Y.; Li, X.; Chen, Y. Building New Countryside in China: A Geographical Perspective. Land Use Policy 2010, 27, 457–470. [Google Scholar] [CrossRef]
- Long, H.; Liu, Y.; Wu, X.; Dong, G. Spatio-Temporal Dynamic Patterns of Farmland and Rural Settlements in Su–Xi–Chang Region: Implications for Building a New Countryside in Coastal China. Land Use Policy 2009, 26, 322–333. [Google Scholar] [CrossRef]
- Lou, J. Tongxiang City in the new situation of rural land comprehensive development of practice and thinking. Zhejiang Land Resour. 2019, 11, 20–23. [Google Scholar] [CrossRef]
- Zheng, X.; Wu, B.; Weston, M.; Zhang, J.; Gan, M.; Zhu, J.; Deng, J.; Wang, K.; Teng, L. Rural Settlement Subdivision by Using Landscape Metrics as Spatial Contextual Information. Remote Sens. 2017, 9, 486. [Google Scholar] [CrossRef]
- Hoeser, T.; Kuenzer, C. Object Detection and Image Segmentation with Deep Learning on Earth Observation Data: A Review-Part I: Evolution and Recent Trends. Remote Sens. 2020, 12, 1667. [Google Scholar] [CrossRef]
- Chen, L.; Li, S.; Bai, Q.; Yang, J.; Jiang, S.; Miao, Y. Review of Image Classification Algorithms Based on Convolutional Neural Networks. Remote Sens. 2021, 13, 4712. [Google Scholar] [CrossRef]
- Huang, X.; Ren, L.; Liu, C.; Wang, Y.; Yu, H.; Schmitt, M.; Hänsch, R.; Sun, X.; Huang, H.; Mayer, H. Urban Building Classification (UBC)—A Dataset for Individual Building Detection and Classification from Satellite Imagery. In Proceedings of the 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), New Orleans, LA, USA, 19–20 June 2022; pp. 1412–1420. [Google Scholar]
- Sun, X.; Huang, X.; Mao, Y.; Sheng, T.; Li, J.; Wang, Z.; Lu, X.; Ma, X.; Tang, D.; Chen, K. GABLE: A First Fine-Grained 3D Building Model of China on a National Scale from Very High Resolution Satellite Imagery. Remote Sens. Environ. 2024, 305, 114057. [Google Scholar] [CrossRef]
- Zheng, X.; Wang, Y.; Gan, M.; Zhang, J.; Teng, L.; Wang, K.; Shen, Z.; Zhang, L. Discrimination of Settlement and Industrial Area Using Landscape Metrics in Rural Region. Remote Sens. 2016, 8, 845. [Google Scholar] [CrossRef]
- Yang, J.; Zhou, Z.; Du, Z.; Xu, Q.; Yin, H.; Liu, R. Rural Construction Land Extraction from High Spatial Resolution Remote Sensing Image Based on SegNet Semantic Segmentation Model. Trans. Chin. Soc. Agric. Eng. 2019, 35, 251–258. [Google Scholar] [CrossRef]
- Ronneberger, O.; Fischer, P.; Brox, T. U-Net: Convolutional Networks for Biomedical Image Segmentation. In Proceedings of the Medical Image Computing and Computer-Assisted Intervention—MICCAI 2015, Munich, Germany, 5–9 October 2015; Navab, N., Hornegger, J., Wells, W.M., Frangi, A.F., Eds.; Springer International Publishing: Cham, Switzerland, 2015; pp. 234–241. [Google Scholar]
- Badrinarayanan, V.; Kendall, A.; Cipolla, R. SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation. IEEE Trans. Pattern Anal. Mach. Intell. 2017, 39, 2481–2495. [Google Scholar] [CrossRef]
- Chen, L.-C.; Zhu, Y.; Papandreou, G.; Schroff, F.; Adam, H. Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation. In Proceedings of the Computer Vision—ECCV 2018, 15th European Conference, Munich, Germany, 8–14 September 2018. [Google Scholar] [CrossRef]
- Wu, P.; Cai, M.; Yi, X.; Wang, G.; Mo, L.; Chola, M.; Kapapa, C. Sweetgum Leaf Spot Image Segmentation and Grading Detection Based on an Improved DeeplabV3+ Network. Forests 2023, 14, 1547. [Google Scholar] [CrossRef]
- Li, J.; Cai, Y.; Li, Q.; Kou, M.; Zhang, T. A Review of Remote Sensing Image Segmentation by Deep Learning Methods. Int. J. Digit. Earth 2024, 17, 2328827. [Google Scholar] [CrossRef]
- Zhang, Z.; Liu, Q.; Wang, Y. Road Extraction by Deep Residual U-Net. IEEE Geosci. Remote Sens. Lett. 2018, 15, 749–753. [Google Scholar] [CrossRef]
- Wang, X.; Hu, Z.; Shi, S.; Hou, M.; Xu, L.; Zhang, X. A Deep Learning Method for Optimizing Semantic Segmentation Accuracy of Remote Sensing Images Based on Improved UNet. Sci. Rep. 2023, 13, 7600. [Google Scholar] [CrossRef] [PubMed]
- Yi, X.; Wang, J.; Wu, P.; Wang, G.; Mo, L.; Lou, X.; Liang, H.; Huang, H.; Lin, E.; Maponde, B.T.; et al. AC-UNet: An Improved UNet-Based Method for Stem and Leaf Segmentation in Betula Luminifera. Front. Plant Sci. 2023, 14, 1268098. [Google Scholar] [CrossRef]
- Shen, J.; Zhang, L.; Yang, L.; Xu, H.; Chen, S.; Ji, J.; Huang, S.; Liang, H.; Dong, C.; Lou, X. Testing a Method Based on an Improved UNet and Skeleton Thinning Algorithm to Obtain Branch Phenotypes of Tall and Valuable Trees Using Abies Beshanzuensis as the Research Sample. Plants 2023, 12, 2444. [Google Scholar] [CrossRef]
- Ye, Z.; Fu, Y.; Gan, M.; Deng, J.; Comber, A.; Wang, K. Building Extraction from Very High Resolution Aerial Imagery Using Joint Attention Deep Neural Network. Remote Sens. 2019, 11, 2970. [Google Scholar] [CrossRef]
- Chen, Z.; Li, D.; Fan, W.; Guan, H.; Wang, C.; Li, J. Self-Attention in Reconstruction Bias U-Net for Semantic Segmentation of Building Rooftops in Optical Remote Sensing Images. Remote Sens. 2021, 13, 2524. [Google Scholar] [CrossRef]
- Ye, Z.; Si, B.; Lin, Y.; Zheng, Q.; Zhou, R.; Huang, L.; Wang, K. Mapping and Discriminating Rural Settlements Using Gaofen-2 Images and a Fully Convolutional Network. Sensors 2020, 20, 6062. [Google Scholar] [CrossRef]
- Chen, L.-C.; Papandreou, G.; Kokkinos, I.; Murphy, K.P.; Yuille, A.L. DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs. IEEE Trans. Pattern Anal. Mach. Intell. 2016, 40, 834–848. [Google Scholar] [CrossRef]
- Li, J.; Huang, Z.-W.; Wang, Y.; Luo, Q. Sea and Land Segmentation of Optical Remote Sensing Images Based on U-Net Optimization. Remote Sens. 2022, 14, 4163. [Google Scholar] [CrossRef]
- Fan, X.; Yan, C.; Fan, J.; Wang, N. Improved U-Net Remote Sensing Classification Algorithm Fusing Attention and Multiscale Features. Remote Sens. 2022, 14, 3591. [Google Scholar] [CrossRef]
- Li, T.; Wang, C.; Wu, F.; Zhang, H.; Tian, S.; Fu, Q.; Xu, L. Built-Up Area Extraction from GF-3 SAR Data Based on a Dual-Attention Transformer Model. Remote Sens. 2022, 14, 4182. [Google Scholar] [CrossRef]
- Roy, A.G.; Navab, N.; Wachinger, C. Concurrent Spatial and Channel ‘Squeeze & Excitation’ in Fully Convolutional Networks. In Proceedings of the Medical Image Computing and Computer Assisted Intervention—MICCAI 2018, Granada, Spain, 16–20 September 2018; Springer: Cham, Switzerland, 2018; pp. 421–429. [Google Scholar]
- Nan, G.; Li, H.; Du, H.; Liu, Z.; Wang, M.; Xu, S. A Semantic Segmentation Method Based on AS-Unet++ for Power Remote Sensing of Images. Sensors 2024, 24, 269. [Google Scholar] [CrossRef]
- Jia, L.; He, X.; Huang, A.; Jia, B.; Wang, X. Highly Efficient Encoder-Decoder Network Based on Multi-Scale Edge Enhancement and Dilated Convolution for LDCT Image Denoising. Signal Image Video Process. 2024, 18, 6081–6091. [Google Scholar] [CrossRef]
- Miao, R.; Meng, G.; Zhou, K.; Li, Y.; Chang, R.; Zhang, G. DASUNet: A Deeply Supervised Change Detection Network Integrating Full-Scale Features. Sci. Rep. 2024, 14, 12464. [Google Scholar] [CrossRef]
- Chen, C.; Fan, L. Scene Segmentation of Remotely Sensed Images with Data Augmentation Using U-Net++. In Proceedings of the 2021 International Conference on Computer Engineering and Artificial Intelligence (ICCEAI), Shanghai, China, 27–29 August 2021; pp. 201–205. [Google Scholar] [CrossRef]
- Zhang, R.; Jiang, P.; Kong, X. Reconstructing Rural Settlements Based on Investigation of Consolidation Potential: Mechanisms and Paths. Land 2024, 13, 354. [Google Scholar] [CrossRef]
- Ran, D.; Hu, Q.; Zhang, Z. Spatial–Temporal Evolution, Impact Mechanisms, and Reclamation Potential of Rural Human Settlements in China. Land 2024, 13, 430. [Google Scholar] [CrossRef]
- Wang, Y.; Zhao, Q.; Wu, Y.; Tian, W.; Zhang, G. SCA-Net: Multiscale Contextual Information Network for Building Extraction Based on High-Resolution Remote Sensing Images. Remote Sens. 2023, 15, 4466. [Google Scholar] [CrossRef]
- Zhou, Z.; Rahman Siddiquee, M.M.; Tajbakhsh, N.; Liang, J. UNet++: A Nested U-Net Architecture for Medical Image Segmentation. In Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support; Stoyanov, D., Taylor, Z., Carneiro, G., Syeda-Mahmood, T., Martel, A., Maier-Hein, L., Tavares, J.M.R.S., Bradley, A., Papa, J.P., Belagiannis, V., et al., Eds.; Lecture Notes in Computer Science; Springer International Publishing: Cham, Switzerland, 2018; Volume 11045, pp. 3–11. ISBN 978-3-030-00888-8. [Google Scholar]
- Shi, Q.; Zhu, J.; Liu, Z.; Guo, H.; Gao, S.; Liu, M.; Liu, Z.; Liu, X. The Last Puzzle of Global Building Footprints—Mapping 280 Million Buildings in East Asia Based on VHR Images. J. Remote Sens. 2024, 4, 0138. [Google Scholar] [CrossRef]
- Wang, P.; Chen, P.; Yuan, Y.; Liu, D.; Huang, Z.; Hou, X.; Cottrell, G. Understanding Convolution for Semantic Segmentation. In Proceedings of the 2018 IEEE Winter Conference on Applications of Computer Vision (WACV), Lake Tahoe, NV, USA, 12–15 March 2017; pp. 1451–1460. [Google Scholar] [CrossRef]
- Hu, L.; Zhou, X.; Ruan, J.; Li, S. ASPP+-LANet: A Multi-Scale Context Extraction Network for Semantic Segmentation of High-Resolution Remote Sensing Images. Remote Sens. 2024, 16, 1036. [Google Scholar] [CrossRef]
- Ding, W.; Huang, H.; Wang, Y. Semantic Segmentation of Remote Sensing Architectural Images Based on GAN and UNet3+ Model. In Proceedings of the PRICAI 2023: Trends in Artificial Intelligence, Jakarta, Indonesia, 15–19 November 2023; Liu, F., Sadanandan, A.A., Pham, D.N., Mursanto, P., Lukose, D., Eds.; Springer Nature: Singapore, 2024; pp. 262–273. [Google Scholar]
- Gao, F.; He, Y.; Wang, J.; Hussain, A.; Zhou, H. Anchor-Free Convolutional Network with Dense Attention Feature Aggregation for Ship Detection in SAR Images. Remote Sens. 2020, 12, 2619. [Google Scholar] [CrossRef]
Characteristic | Dispersed Rural Settlements (DRSs) | Clustered Rural Settlements (CRSs) |
---|---|---|
Composition | The main building, affiliated facilities (livestock pens and barns), gardens, woodlands, etc. | Neatly arranged standardized houses surrounded by less vegetation cover |
Spatial morphology | Buildings are scattered and have various orientations, no distinct distribution pattern, low building density | Homogeneous building types, moderate or dense building density, orderly arrangement in space |
Location | Close to country footpaths, rivers, and streams for transportation and water supply | Close to main roads for easy transportation |
Stage | Output Size | Template |
---|---|---|
Input | 3 × 256 × 256 | |
Conv1 | 64 × 256 × 256 | |
Maxpool | 64 × 128 × 128 | |
Conv2 | 128 × 128 × 128 | |
Maxpool | 128 × 64 × 64 | |
Conv3 | 256 × 64 × 64 | |
Maxpool | 256 × 32 × 32 | |
Conv4 | 512 × 32 × 32 | |
Maxpool | 512 × 16 × 16 | |
Conv5 | 512 × 16 × 16 | |
Up Layer1 | 512 × 16 × 16 | Upsample BatchNormal |
Up Layer2 | 512 × 32 × 32 | Upsample BatchNormal |
Up Layer3 | 256 × 64 × 64 | Upsample BatchNormal |
Up Layer4 | 128 × 128 × 128 | Upsample BatchNormal |
Up Layer5 | 64 × 256 × 256 | Upsample BatchNormal |
Segmentation Layer | 3 × 256 × 256 |
Model | Precision (%) | Recall (%) | F1 (%) | MioU (%) |
---|---|---|---|---|
UNet | 91.38 | 92.35 | 91.86 | 85.43 |
ASCEND-UNet | 96.05 | 95.15 | 95.59 | 91.71 |
DeepLabv3+ | 87.90 | 86.02 | 86.94 | 78.06 |
PSPNet | 83.02 | 74.99 | 78.63 | 67.32 |
SegNet | 90.37 | 77.54 | 83.46 | 71.62 |
Model | Traning Time | Inference Time |
---|---|---|
UNet | ~6.5 h | 0 m 26 s |
ASCEND-UNet | ~7.5 h | 0 m 30 s |
Deeplabv3+ | ~4.6 h | 0 m 25 s |
PSPNet | ~5.2 h | 0 m 27 s |
SegNet | ~4.8 h | 0 m 28 s |
Model | Precision (%) | Recall (%) | F1-Score (%) | MioU (%) |
---|---|---|---|---|
UNet | 91.38 | 92.35 | 91.86 | 85.43 |
UNet + scSE | 95.40 | 94.54 | 94.97 | 90.61 |
UNet + ASPP | 94.99 | 94.54 | 94.76 | 90.25 |
UNet + ASPP + HDC | 95.88 | 94.87 | 95.27 | 91.32 |
UNet + scSE + ASPP | 95.03 | 94.50 | 94.76 | 90.25 |
UNet + HDC | 95.98 | 94.86 | 95.41 | 91.39 |
ASCEND-UNet | 96.05 | 95.15 | 95.59 | 91.71 |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2024 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Zheng, X.; Pu, S.; Xue, X. ASCEND-UNet: An Improved UNet Configuration Optimized for Rural Settlements Mapping. Sensors 2024, 24, 5453. https://doi.org/10.3390/s24175453
Zheng X, Pu S, Xue X. ASCEND-UNet: An Improved UNet Configuration Optimized for Rural Settlements Mapping. Sensors. 2024; 24(17):5453. https://doi.org/10.3390/s24175453
Chicago/Turabian StyleZheng, Xinyu, Shengwei Pu, and Xingyu Xue. 2024. "ASCEND-UNet: An Improved UNet Configuration Optimized for Rural Settlements Mapping" Sensors 24, no. 17: 5453. https://doi.org/10.3390/s24175453