This is a test project to assess and show how we could extract relevant data about damage caused by an earthquake on villages and people.
The aims of this project are three-fold:
- To categorize on-the-ground photos taken in the aftermath of an earthquake
- To extract data on the damage caused by the earthquake on buildings, infrastructure, and displacements, as well as data about displaced persons’ situation
- To use this process to develop a protocol for rapid earthquake response using GeoTag-X and PyBossa
The photographs collected in this project have been taken on the ground in the aftermath of the April 2015 Nepal Earthquake, and uploaded to a web-server with UN-ASIGN. UN-ASIGN is a free crowd-sourcing app developed by UNOSAT that automatically georeferences photos and sends them to a central server. These photos can serve as an important media source for GeoTag-X and are the initial data source for this project.
To explore how GeoTag-X could support relief efforts in future during disaster events, it would be helpful to categorize the UN-ASIGN photographs we have collected by assessing what the photographs depict and what types of damage are visible, and to extract that data from them. We will then assess these results and determine how GeoTag-X can be deployed for the next earthquake.
In this project, volunteers are asked to answer questions for each photo as best as they can. To help in the assessment, we are asking questions based on UN OCHA’s (United Nations Office for the Coordination of Humanitarian Affairs) assessment forms for Nepal. In particular, we want you to answer questions like:
- Do you see housing, infrastructure, or local social infrastructure in this photo?
- Do you see displaced population or people requiring assistance in this photo?
- Do you see any possible source of drinking water in this photo?
- Do you see any sanitation or hygiene issue in this photo?
- Do you see crops or livestock in this photo?
How does it work?
To contribute to the project, you only need to answer the questions as best you can. In order to better perform each task, we highly recommend taking the tutorial before starting so you know what to look for.
If you need some pointers, you can click on the help for each question anytime you wish.
Don't worry if you are not sure about your answer, each photo is presented to at least 30 people and the final answers are taken from the majority response. So there is no problem if you get a couple wrong.
Just do your best – and don’t forget that you can always take a look at the learning tools we provide (tutorial, help boxes for each question, and source link under the photo)
Sometimes it will be difficult to see what is in the photo - if you are not sure, don't be afraid to select "I DON'T KNOW" – this is also a valid and useful answer!
You can also select “IMAGE NOT CLEAR” if you think that the photo is not clear enough for you to answer the question, e.g. when the photo is too blurred.
Don't worry if you find you have a lot of photos that are not relevant, sometimes you might go through several photos in a row of buildings that might be hard to classify, or photos that are not showing people, etc. Photos in this project are covering many different topics like buildings, shelter, people, and so on. It is completely fine that you are not able to assess all these elements in the same photo.
What are we doing with the results?
Results generated by your contributions to this project could potentially be deployed in future during a disaster situation, when another earthquake will strike, and for disaster preparedness.
The same results are also meant to complement other sources of data informing the response and recovery effort for the 2015 Nepal Earthquake, such as satellite imagery analysis and field assessment.
Project co-developed by Valentina Rigamonti with UN OCHA.