Omega Alpha Omega| Spot That Fire V3.0

Spot That Fire V3.0

Recent wildfires worldwide have demonstrated the importance of rapid wildfire detection, mitigation, and community impact assessment analysis. Your challenge is to develop and/or augment an existing application to detect, predict, and assess the economic impacts from actual or potential wildfires by leveraging high-frequency data from a new generation of geostationary satellites, data from polar-orbiting environmental satellites, and other open-source datasets.

Eye in the sky

Summary

During emergencies, it is common for people to lose composure and hence make wrong decisions which may prove to be fatal. Our project 'Eye in the sky' works to mitigate the loss of human lives as during such scenarios it'd be best to have the opinion and instructions determined by the experts in the field ASAP. Our project consists of a web app and an android app. The web app uses computer vision to analyze the satellite imagery to determine if there is any potential inferno near any human settlement. If so, we can send alert via the web app and spread the alert message(s) using the android app to the masses along with any updates and/or set of instructions that are to be followed.

How We Addressed This Challenge

We developed a web app and an app. The web app uses the computer vision algorithm to analyze if there's any potential inferno near any human settlement. The deployed model has an accuracy of 100% on validation set whereas 97.5% on test set. If the output comes out positive, then entering the coordinates of the place we can issue an alert.

That alert is then received by the experts/officials that then decide upon the set of certain actions (like evacuating from some part of the city 'X' to some part of the city 'Y') that needs to be followed by the civilians. If the civilians are having certain issues they may directly call or message the experts/officials too. The app also has a donation feature as well.

We hope to mitigate the loss of human lives through our project. In future enhancements, we also hope to use this alert system not only for fire but for different disasters too. (Like the Bhopal Gas tragedy of 1984)

How We Developed This Project

Our team is from Bhopal and as mentioned earlier Bhopal was the site of the "Worst industrial disaster". Our parents were witness of that horrific incident. We thought that if people knew the correct course of action to take back then, it'd save so many lives.

We held team conferences to come up with our solution and decided upon the technologies that we'll need to incorporate to make our project. We used Python, Django, Tensorflow, Keras, CSS, Java Script, flutter, bootstrap and SQLite to make our project.

Our team faced various problems from poor performance of the model to many bugs in the app. At one point we were not even able to deploy the model on our web app. But at the end we were able to make the ends meet.

How We Used Space Agency Data in This Project

We used the images from the HIMAWARI geostationary satellite in the spot that fire resource in order to train our model. We took around 200 images in the .hdf format in 500m grid for the positive class as well as 200 satellite images from google images for the negative class.

Project Demo

https://www.youtube.com/watch?v=AutEnykdbNY&feature=youtu.be

Project Code
Data & Resources

https://data.nas.nasa.gov/geonex/data.php?dir=/geonexdata/HIMAWARI8/GEONEX-L1G

https://firms.modaps.eosdis.nasa.gov/map/

Tags
#artificial_intelligence, #computer_vision, #software, #web_app, #android_app, #bushfire, #wildfire
Judging
This project was submitted for consideration during the Space Apps Judging process.