Deep learning for integration of Satellite images and google StreetView for mapping informal settlements (slums)
Suggested elective courses: Advanced Image Analysis, Computer vision related courses
For this topic, the student must have good programing skills.
Mapping and monitoring slums is essential to ensure the sustainable development of the cities to ensure a quality of life for people. Despite its importance, the coarser resolution of freely available remote sensing images makes it represents a great methodological challenge, bringing the need of including auxiliary data. In that sense, Google StreetView offer a huge database of ground level images that can be used to improve the efficiency of these mapping tasks. Several works have demonstrated the potential of Deep Learning methods to combine Remote Sensing data with ground level images to perform mapping tasks. Therefore, this topic focus on the development of an efficient Deep Learning model able to combine Remote Sensing data with Google StreetView images for mapping slums.
The student will initially revise the state of the art of semantic segmentation models, feature transferability and multimodal deep learning, aiming to find the most efficient and accurate models. The next step will be the comparison between those methods, and their optimization, aiming in this way to achieve a better model with a good computational cost and accuracy.
Najmi A, Gevaert CM, Kohli D, Kuffer M, Pratomo J. Integrating Remote Sensing and Street View Imagery for Mapping Slums. ISPRS International Journal of Geo-Information. 2022; 11(12):631. https://doi.org/10.3390/ijgi11120631
R. Fan, J. Li, F. Li, W. Han and L. Wang, "Multilevel Spatial-Channel Feature Fusion Network for Urban Village Classification by Fusing Satellite and Streetview Images," in IEEE Transactions on Geoscience and Remote Sensing, vol. 60, pp. 1-13, 2022, Art no. 5630813, doi: 10.1109/TGRS.2022.3208166.
Chen, B., Feng, Q., Niu, B., Yan, F., Gao, B., Yang, J., ... & Liu, J. (2022). Multi-modal fusion of satellite and street-view images for urban village classification based on a dual-branch deep neural network. International Journal of Applied Earth Observation and Geoinformation, 109, 102794.