What is Tomnod?
Out in the real world, one of the best ways to find things or people that are lost, whether lost keys or lost kids, is to recruit help to find them. In a nutshell, that’s what Tomnod is. In fact, it has been called a “virtual search party”. Tomnod is a crowdsourcing application that lets ordinary people all over the world look at satellite imagery and answer simple questions about it.
Crowdsourcing is a relatively new term for getting people to help in a task. Rather than hiring or recruiting people to work on a task, the task is put out there for the public to assist. This allows for a lot more eyes on a given task, allowing for faster results and greater accuracy. In Tomnod, these tasks are called campaigns.
One of the most prominent use cases for Tomnod was during the disappearance of the Malaysian Aircraft flight 229. DigitalGlobe’s Tomnod site had over six million views of its Tomnod imagery, as the crowd was leveraged to tag anything relevant to an airliner crash. Try to imagine it, people all over the world, combing through photographs of the ocean, trying to locate the missing plane. That’s what harnessing the crowd can do for you.
Campaigns take two forms: discovery campaigns, where users identify features in a given image, and labeling campaigns, where users can label or classify features in imagery. Labeling campaigns can also be used to validate or further refine objects found in a previous discovery campaign.
In discovery campaigns, a user is given a random satellite image from a given area and asked to identify specific features. Currently, there’s an active discovery campaign looking for Weddell seals in Antarctica, as well as one looking for settlements in South Sudan. Check out tomnod.com to participate.
Labelling campaigns ask the public to classify a marked area in an image from a short list of options. This harnesses the crowd to really refine what’s in each image.
The ability of the crowd can be harnessed in many ways. While finding and classifying objects using the crowd is an awesome capability, that’s not always the end of the story. In the next post, we’ll discuss how Tomnod can be used with DigitalGlobe’s machine learning software to create even more powerful workflows.