I have a project about "Flood Detection", I'm trying to copy the method of "Deep-Learning Integration of CNN–Transformer and U-Net for Bi-Temporal SAR Flash-Flood Detection" which is published paper 2025 by Abbas Mohammed Noori.
what I did until now : I downloaded SAR Images from Copernicus: all from Sentinel-1, IW, Level1, VH, 10m. SAR Image for Pre-flood: 17 Nov 2022. SAR Image for Post-flood: 29 Nov 2022.
I obtained two GroundTruths from "GFM" Global Flood Monitoring, one for the pre-flood period on 17 Nov 2022. And the other one for Post-flood: 29 Nov 2022.
I performed the preprocessing for SAR images using SNAP.
I will share with you: 1. SAR images. 2. Ground truths. 3. Code for doing dataset patches, Training, Testing. 4. The paper that I'm trying to follow it's method. 5. Jeddah city DEM. 6. Jeddah city Shapefile.
What I need: I need a Prediction Flood Map with high Accuracy.
Please fix my code!
The due date is Thursday!
Please follow the instructions in the paper!
I need someone with extensive experience in Deep Learning for satellite images.
Web & Social Data Entry Category: Data Collection, Data Entry, Data Management, Excel, Google Sheets, Social Media Management, Web Scraping, Web Search Budget: ₹750 - ₹1250 INR
