Remotely sensed data can provide information about conditions on the ground that may affect food supply chains and food security during pandemics. Your challenge is to demonstrate the agricultural socio-economic impacts from COVID-19.
Our project is a software program that detects what is what in a satellite image based on color and labels them in the following categories: water body, cultivated land, unused land, wheat, buildings and other man-made objects, forest, solar panels, clouds. We hope that our software will help people see by themselves the impact of the pandemic on land use.
We, ERPCTA Kagomi, have developed a software that detects land use from satellite images and labels them in the following categories: water body, cultivated land, unused land, wheat, buildings and other man-made objects, forest, solar panels, clouds.
Our program helps people experiment and view the differences made by the pandemic by themselves, we believe that people learn better what they discover themselves, thus we made a simple software that they can use and can be integrated into the dashboard.
Our program works by dissecting the image into lots of smaller 30x30 pixel images, afterwards the software analyzes each image for the predominant colour, each colour is linked with one of the categories, the program then compares the colours it got from the images to the ones from the database we have created and assigns each image a new colour that represents our programs approximations and reconstructs the full image.
We hope to give people a easy tool to view for themselves the land use from a satellite image or map, in hopes of helping them see the effects of the pandemic.
Our journey was definitely not an easy one, from the start we have struggled to get the data from SentinelHub in hopes of getting the satellite images needed to test our software and develop it.Then we have created the code in Python 3, we had errors after errors as coding works, but we have passed all the challenges we have met.We have learned a lot through this competition, from using the SentinelHub to working with images.We have a computer with Linux and a computer with dual boot Linux and Windows.
We were inspired to choose the agricultural challange from our country's past as "The Granary Of Europe", or at least that's what our history textbooks say.
We have used satellite images from the Sentinel 2 L1C satellite for the development of our software, also the software was inspired by JAXA's Advanced Land Observing Satellite or ALOS System.
The solution can be integrated into the Earth Observing Dashboard by having the Python script running on the server and processing images, and a web page that shows the output image (the result). Source code and documentation: https://github.com/gabi200/eodashhack
SentinelHub: Sentinel 2 L1C - This is where we got our satellite images from.
JAXA ALOS - this is where we got inspiration to do what we did from.
#agriculture #interactive
This project has been submitted for consideration during the Judging process.