ICD: VHR-Oriented Interactive Change-Detection Algorithm
Abstract
:1. Introduction
2. Materials and Methods
2.1. Structure of the Interactive Change Detection Network
2.2. Automatic Generation of Clicks
Algorithm 1 Automatic generation algorithm of clicks. |
Input: The latest generation model, two-phase training samples A, B and the corresponding label; Output: Positive Clicks and Negative Clicks for current model training results. |
1: Use the model to predict the training samples, and obtain the prediction results; 2: Compare the prediction results with the labels to determine the areas that need to be modified by positive and negative clicks. These areas are defined as follows:
4: Use connected components to divide the pixels of False Negative and False Positive into different clusters, and obtain the cluster sets F.N.Set and F.P.Set, respectively; 5: Determine the first N large areas maxN(F.N.Set) and maxN(F.P.Set) of F.N.Set and F.P.Set according to the number of pixels; 6: Take the centroid in maxN(F.N.Set) and maxN(F.P.Set) as Positive Clicks and Negative Clicks, respectively; |
7: Update the Positive Clicks and Negative Clicks of the training samples; 8: Return Positive Clicks and Negative Clicks. |
2.3. The Evaluation Index
- Correction range of positive clicks
- 2.
- Negative click-correction range
2.4. Loss Function
3. Results
3.1. Datasets
3.2. Accuracy Test
4. Discussion
4.1. Discussion of the Experiment
4.2. Application Scenario
4.3. Deficiency and Future Improvement Direction
Author Contributions
Funding
Data Availability Statement
Conflicts of Interest
References
- Ashbindu, S. Review Article Digital Change Detection Techniques Using Remotely-Sensed Data. Int. J. Remote Sens. 1988, 10, 989–1003. [Google Scholar]
- Haigang, S.; Wenqing, F.; Wenzhuo, L.; Kaimin, S.; Chuan, X. Review of Change Detection Methods for Muti-temporal Rempte Sensing Imagery. Geomat. Inf. Sci. Wuhan Univ. 2018, 43, 1885–1898. [Google Scholar]
- Hussain, M.; Chen, D.; Cheng, A.; Wei, H.; Stanley, D. Change Detection from Remotely Sensed Images:From Pixel-Based to Object-Based Approaches. ISPRS J. Photogramm. Remote Sens. 2013, 80, 91–106. [Google Scholar] [CrossRef]
- Karantzalos, K. Recent Advances on 2D and 3D Change Detection in Urban Environments from Remote Sensing Data. Comput. Approaches Urban Environ. 2014, 13, 237–272. [Google Scholar]
- Lu, D.; Mausel, P.; Brondizio, E.; Moran, E. Change Detection Techniques. Int. J. Remote Sens. 2004, 25, 2365–2401. [Google Scholar] [CrossRef]
- Zhigao, Y.; Qianqing, Q.; Qifeng, Z. Change Detection in High Spatial Resolution Images Based on Support Vector Machine. In Proceedings of the IEEE International Conference on Geoscience and Remote Sensing Symposium 2006, Denver, CO, USA, 31 July–4 August 2006. [Google Scholar] [CrossRef]
- Huang, X.; Xie, Y.; Wei, J.; Fu, M.; Lv, L.L.; Zhang, L.L. Automatic Recognition of Desertification Information Based on the Pattern of Change Detection-CART Decision Tree. J. Catastrophol. 2017, 31, 36–42. [Google Scholar]
- Seo, D.K.; Yong, H.K.; Yang, D.E.; Park, W.Y.; Park, H.C. Generation of Radiometric, Phenological Normalized Image Based on Random Forest Regression for Change Detection. Remote Sens. 2017, 9, 1163. [Google Scholar] [CrossRef]
- Lu, J.; Ming, L.; Peng, Z.; Yan, W.; Huahui, Z. SAR Image Change Detection Based on Multiple Kernel k-Means Clustering with Local-Neighborhood Information. IEEE Geosci. Remote Sens. Lett. 2016, 13, 856–860. [Google Scholar]
- Lv, H.; Lu, H.; Mou, I. Learning a Transferable Change Rule from a Recurrent Neural Network for Land Cover Change Detection. Remote Sens. 2016, 8, 506. [Google Scholar]
- Lu, J.; Ming, L.; Peng, Z.; Wu, Y. SAR Image Change Detection Based on Correlation Kernel and Multistage Extreme Learning Machine. IEEE Trans. Geosci. Remote Sens. 2016, 54, 5993–6006. [Google Scholar]
- Guo, C.; Yazhou, L.; Yanfeng, S. Automatic Change Detection in Remote Sensing Images Using Level Set Method with Neighborhood Constraints. J. Appl. Remote Sens. 2014, 8, 83–678. [Google Scholar]
- Ming, H.; Hua, Z.; WenZhong, S.; Kazhong, D. Unsupervised Change Detection Using Fuzzy-Means and MRF from Remotely Sensed Images. Remote Sens. Lett. 2013, 4, 1185–1194. [Google Scholar]
- Licun, Z.; Guo, C.; Yupeng, L.; Yanfeng, S. Change Detection Based on Conditional Random Field with Region Connection Constraints in High-Resolution Remote Sensing Images. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. 2016, 9, 3478–3488. [Google Scholar]
- Khelifi, L.; Mignotte, M. Deep Learning for Change Detection in Remote Sensing Images: Comprehensive Review and Meta-Analysis. IEEE Access 2020, 8, 126385–126400. [Google Scholar] [CrossRef]
- Hao, C.; Zhenwei, S. A Spatial-Temporal Attention-Based Method and a New Dataset for Remote Sensing Image Change Detection. Remote Sens. 2020, 12, 1662. [Google Scholar]
- Jie, C.; Ziyuan, Y.; Li, C.; Jian, P.; Haozhe, H.; Jiawei, Z.; Yu, L.; Haifeng, L. DASNet: Dual attentive fully convolutional siamese networks for change detection of high resolution satellite images. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. 2020, 14, 1194–1206. [Google Scholar]
- Ken, S.; Mikiya, S.; Weimin, W. Weakly Supervised Silhouette-based Semantic Scene Change Detection. In Proceedings of the IEEE International Conference on Robotics and Automation (ICRA), Paris, France, 31 May–31 August 2020; pp. 6861–6867. [Google Scholar]
- Mengxi, L.; Qian, S.; Andrea, M.; Da, H.; Xiaoping, L.; Liangpei, Z. Super-resolution-based Change Detection Network with Stacked Attention Module for Images with Different Resolutions. IEEE Trans. Geosci. Remote Sens. 2021, 60, 4403718. [Google Scholar]
- Lin, W.; Haiyan, L. Interactive Segmentation Algorithm for Bio-medicine Image. Comput. Eng. 2010, 36, 208–212. [Google Scholar]
- Wei, H.; Qingwu, H.; Mingyao, A. Interactive Remote Sensing Image Segmentation Based on Multiple Star Prior and Graph Cuts. Remote Sens. Inf. 2016, 31, 19–23. [Google Scholar]
- Sofiiuk, K.; Petrov, I.A.; Konushin, A. Reviving Iterative Training with Mask Guidance for Interactive Segmentation. arXiv 2021. [Google Scholar] [CrossRef]
- Huayue, Z.; Shunli, Z.; Li, Z. Interactive Target Segmentation Algorithm Based on Two-Stage Network. Comput. Eng. 2021, 47, 300–306. [Google Scholar]
- Castrejon, L.; Kundu, K.; Urtasun, R.; Fidler, S. Annotating Object Instances with a Polygon-RNN. In Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2017, Honolulu, HI, USA, 21–26 July 2017. [Google Scholar]
- Sahbi, H.; Deschamps, S.; Stoian, A. Active learning for interactive satellite image change detection. arXiv 2021. [Google Scholar] [CrossRef]
- Hui, R.; Huai, Y.; Pingping, H.; Wen, Y. Interactive Change Detection Using High Resolution Remote Sensing Images Based on Active Learning with Gaussian Processes. ISPRS Ann. Photogramm. Remote Sens. Spat. Inf. Sci. 2016, 3, 141. [Google Scholar]
- Hichri, H.; Bazi, Y.; Alajlan, N.; Malek, S. Interactive Segmentation for Change Detection in Multispectral Remote-Sensing Images. IEEE Geosci. Remote Sens. Lett. 2013, 10, 298–302. [Google Scholar] [CrossRef]
- Saux, B.L.; Randrianarivo, H. Urban Change Detection In SAR Images By Interactive Learning. In Proceedings of the 2013 IEEE International Geoscience and Remote Sensing Symposium, Melbourne, VIC, Australia, 21–26 July 2013. [Google Scholar] [CrossRef]
- Mahadevan, S.; Voigtlaender, P.; Leibe, B. Iteratively trained interactive segmentation. arXiv 2018. [Google Scholar] [CrossRef]
- Junhao, L.; Yunchao, W.; Wei, X.; Ong, S.-H.; Feng, J. Regional Interactive Image Segmentation Networks. In Proceedings of the 2017 IEEE International Conference on Computer Vision (ICCV) 2017, Venice, Italy, 22–29 October 2017; pp. 2746–2754. [Google Scholar] [CrossRef]
- Honglin, W.; Lichen, J.; Fangfang, X. Interactive Segmentation Technique and Decision-level Fusion Based Change Detection for SAR Images. Acta Geod. Cart. Sin. 2012, 41, 74–80. [Google Scholar]
- Xinying, H.; Jiazhong, W.; Chenxia, S.; Chang, S. Edge detection method based on mathematical morphology and canny algorithm. J. Comput. Appl. 2008, 8, 477–478. [Google Scholar]
- Shunping, J.; Shiqing, W. Building extraction via convolutional neural networks from an open remote sensing building dataset. Acta Geod. Cartogr. Sin. 2019, 48, 448–459. [Google Scholar]
Basic Information | Type | Value |
---|---|---|
image information | sample size | 256 × 256 |
resolution | 0.2 m | |
change information | total change instances | 2421 |
training sets | 4960 | |
verification sets | 1417 | |
test sets | 709 |
Basic Information | Type | Value |
---|---|---|
image information | sample size | 256 × 256 |
resolution | 0.5 m | |
change information | total change instances | 31,333 |
training sets | 7942 | |
verification sets | 2269 | |
test sets | 1135 |
Net | Dataset | Clicks | Accuracy | Precision | Recall | F1 | CR_Positive | CR_Negative |
---|---|---|---|---|---|---|---|---|
Gradient_Res | WHU | 0 | 0.965 | 0.852 | 0.902 | 0.877 | ||
1 | 0.976 | 0.882 | 0.953 | 0.916 | 0.776 | 0.531 | ||
5 | 0.976 | 0.874 | 0.968 | 0.919 | 0.779 | 0.546 | ||
10 | 0.977 | 0.878 | 0.965 | 0.920 | 0.779 | 0.546 | ||
Disc_Res | WHU | 0 | 0.965 | 0.853 | 0.900 | 0.876 | ||
1 | 0.971 | 0.864 | 0.941 | 0.901 | 0.896 | 0.493 | ||
5 | 0.974 | 0.871 | 0.952 | 0.910 | 0.876 | 0.510 | ||
10 | 0.974 | 0.872 | 0.952 | 0.910 | 0.876 | 0.507 | ||
Gradient_No_Res | WHU | 0 | 0.965 | 0.855 | 0.897 | 0.876 | ||
1 | 0.966 | 0.862 | 0.902 | 0.882 | 0.313 | 0.332 | ||
5 | 0.969 | 0.872 | 0.906 | 0.889 | 0.351 | 0.357 | ||
10 | 0.969 | 0.874 | 0.907 | 0.890 | 0.352 | 0.355 | ||
Gradient_Res | LEVIR-CD | 0 | 0.976 | 0.919 | 0.820 | 0.870 | ||
1 | 0.980 | 0.921 | 0.869 | 0.894 | 0.794 | 0.633 | ||
5 | 0.982 | 0.893 | 0.927 | 0.909 | 0.814 | 0.654 | ||
10 | 0.983 | 0.905 | 0.919 | 0.912 | 0.815 | 0.655 | ||
Disc_Res | LEVIR-CD | 0 | 0.976 | 0.904 | 0.842 | 0.873 | ||
1 | 0.981 | 0.902 | 0.876 | 0.889 | 0.807 | 0.518 | ||
5 | 0.981 | 0.912 | 0.886 | 0.899 | 0.832 | 0.533 | ||
10 | 0.981 | 0.903 | 0.901 | 0.902 | 0.835 | 0.532 | ||
Gradient_No_Res | LEVIR-CD | 0 | 0.976 | 0.873 | 0.871 | 0.872 | ||
1 | 0.978 | 0.905 | 0.851 | 0.875 | 0.447 | 0.465 | ||
5 | 0.980 | 0.880 | 0.881 | 0.881 | 0.490 | 0.476 | ||
10 | 0.980 | 0.880 | 0.882 | 0.881 | 0.495 | 0.476 |
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |
© 2022 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
Jiang, Z.; Zhou, X.; Cao, W.; Sun, Z.; Wu, C. ICD: VHR-Oriented Interactive Change-Detection Algorithm. ISPRS Int. J. Geo-Inf. 2022, 11, 503. https://doi.org/10.3390/ijgi11100503
Jiang Z, Zhou X, Cao W, Sun Z, Wu C. ICD: VHR-Oriented Interactive Change-Detection Algorithm. ISPRS International Journal of Geo-Information. 2022; 11(10):503. https://doi.org/10.3390/ijgi11100503
Chicago/Turabian StyleJiang, Zhuoran, Xinxin Zhou, Wei Cao, Zaihong Sun, and Changbin Wu. 2022. "ICD: VHR-Oriented Interactive Change-Detection Algorithm" ISPRS International Journal of Geo-Information 11, no. 10: 503. https://doi.org/10.3390/ijgi11100503
APA StyleJiang, Z., Zhou, X., Cao, W., Sun, Z., & Wu, C. (2022). ICD: VHR-Oriented Interactive Change-Detection Algorithm. ISPRS International Journal of Geo-Information, 11(10), 503. https://doi.org/10.3390/ijgi11100503