Team Technocrats

https://github.com/SamaUddinEraj/Nasa-Space-App-Challenge-22

https://drive.google.com/file/d/1omwqBZnOJ0RAVj5FoSUaw5hNDXO_uMar/view?usp=sharing

Flood mapping is a technique to analyze the extent of flood water and identities its spatial distribution over the land area. Early and accurate detection of the geographic areas affected by the flood is crucial. It helps in decision-making for the rescue team by understanding the severity of the situation and making a rescue plan. Sentinel, Landsat, and MODIS are satellites, which provide high-resolution images. We have used the data from the sentinel -2 satellite. With our app, most individuals will be within the protection boundary as they will be aware of the forecasting. 

 Necessary data of shelters, proper division of relief, funds, and emergency alerts will be provided.

 Proposed an automatic flood detection technique by fusing a modified water index technique and deep CNN model (VGG16). • Modified water index is the combination of two indices, Green/SWIR and Blue/NIR. • Results shows that automatic flood detection is possible when an appropriate water index technique is being used. • Future Work : Use of more advance networks in Deep CNN like DenseNet, or GAN • Multi Band Analysis by including all bands of Sentinel-1. 

• Google Earth
• Sentinal- 1

It has been a great journey and our team got to learn a lot of new things.

• Google Earth Engine
• Github

#flood #rescue #googleearth #sentinel1 #app