3.6 million participants have helped look at satellite maps in the hunt for MH370
Published on Mar 20, 2014 2:37 PM
Satellite company DigitalGlobe, which set up a crowdsourcing effort to help look for Malaysia Airlines flight MH370, confirmed on Thursday that its images played a part in the Australian government's announcement that they were following a lead in the sea off its coastline.
“We can confirm that DigitalGlobe has provided imagery to search officials in Australia, and we have been informed by an Australian government official that it was our imagery Prime Minister Abbott referred to in his recent comments. We do not have any additional information at this time, as relevant search parties follow-up on this important lead. However, we will continue to cooperate with authorities to provide any and all information at our disposal to assist the search," Mr Takayuki Odawara, vice president, head of APAC (Asia Pacific) Commercial Sales with DigitalGlobe, told The Straits Times.
The company created Tomnod, http://www.tomnod.com/nod/, a platform for volunteers to scour its archive of satellite images to help hunt for possible signs of the plane.
There have been some 3.6 million participants and 426 million map views since the company announced its initiative on March 10. Mr Odawara added that with so many pairs of eyes looking at the company's database of images, "every pixel has been looked at by live human eyes at least 30 times".
This crowdsourcing effort supplements a hunt that now encompasses 25 nations and 2.24 million nautical square miles.
Mr Odaware added: "Our analysts are monitoring all the news, reports and rumour about the possible search locations and updating the tasking directions for our five satellites accordingly. As important locations change, our satellites will follow. So far, the images at Tomnod platform have covered Gulf of Thailand, Malacca Strait and part of Indian Ocean."
So far, Mr Odawara said, volunteers have tagged more than 6.7 million features. The tagging can be seen on the map on Tomnod. With so much information coming in, the platform uses an algorithm, CrowdRank, to check for overlaps in the tags by volunteers. These tags are then examined again by the company's analysts who will then shortlist further the possibilities and share the information with the authorities.
To continue reading, log in if you are a subscriber
If you are not a subscriber, you can get instant, unlimited access here