recentpopularlog in

satelliteimagery

« earlier   
Concrete and coral: Tracking expansion in the South China Sea - Data Journalism Awards
The data gathering process was the most innovative part of this piece. Hundreds of satellite images taken over years were digitally scrutinised by running them through an algorithm to identify, count, trace and analyse every single building structure in a given area. While most media had focused on land reclamation on the islands, Reuters showed the extent and intensity of building construction taking place above ground. Reuters used the algorithm to collate an exclusive dataset, which readers could breakdown by country and island group.
satelliteimagery  journalism  algorithms  reuters  DJA 
may 2019 by fcoel
Tracking China’s Muslim Gulag - Data Journalism Awards
, the team shed light on the sensitive and secretive area of Xinjiang, where China stands accused of detaining hundreds of thousands of Muslims in high-security compounds. Hundreds of satellite images taken over years were digitally scrutinised. Running them through an algorithm allowed the team to identify, count, trace and analyse every single building structure in a given area. Reuters used the data to publish a ground-breaking special report
dja  china  algorithms  reuters  sea  coral  concrete  satelliteimagery 
may 2019 by fcoel
AP Explore: Seafood from slaves | Associated Press
behind the scenes from AP report The Future of Augmented Journalism: // In one high-profile case, AP used satellite imagery from a company called
Digital Globe to secure high-resolution images of sea vessels in Southeast Asia
to document critical evidence for an investigative project on abuses in the
seafood industry that won the Pulitzer Prize for Public Service in 2016.
Digital Globe’s computer-vision algorithms worked to reorient its
satellite-based cameras to shoot the optimal and necessary images that
ultimately provided a point of reference in the investigation beyond the
reach of AP’s reporting team.
This sort of image recognition works using a type of machine learning called
“neural networks,” according to Matt Zeiler, the founder of an image- and
video-recognition platform called Clarifai. These neural-network algorithms
work by mimicking the way humans are understood to perceive images.
“Nobody really knows exactly how the brain operates, but we do know the
visual cortex processes the input to our eyes in multiple layers,” Zeiler said.
“Our eyes then group regions of images together and keep only the positive
elements. Computer-vision algorithms use mathematical models to replicate
this process through their own layering of images.”
satelliteimagery  ap  journalism  vision  slaves  pulitzer  asia  fish  sea  tools  digitalglobee  imagerecognition  neuralnetworks 
may 2019 by fcoel
Tracking China’s Muslim Gulag
, the team shed light on the sensitive and secretive area of Xinjiang, where China stands accused of detaining hundreds of thousands of Muslims in high-security compounds. Hundreds of satellite images taken over years were digitally scrutinised. Running them through an algorithm allowed the team to identify, count, trace and analyse every single building structure in a given area. Reuters used the data to publish a ground-breaking special report
journalism  reuters  china  muslim  camps  satelliteimagery  dja  algorithms 
may 2019 by fcoel
Concrete and coral
The data gathering process was the most innovative part of this piece. Hundreds of satellite images taken over years were digitally scrutinised by running them through an algorithm to identify, count, trace and analyse every single building structure in a given area. While most media had focused on land reclamation on the islands, Reuters showed the extent and intensity of building construction taking place above ground. Reuters used the algorithm to collate an exclusive dataset, which readers could breakdown by country and island group.
satelliteimagery  journalism  algorithms  reuters  DJA 
may 2019 by fcoel
How AI Helps Human Rights Watch Investigate from the Sky | NVIDIA Blog
But it takes an expert eye — or a neural network — to tell the difference between smoke plumes and puffy white clouds.

“Most of the time, it’s my eyes that are doing the analysis,” Lyons said. “The DGX immediately gives us the ability to scale.”

A deployed deep learning model that analyzes satellite or social media data could one day identify potential human rights abuses automatically from text and images and alert Human Rights Watch and humanitarian agencies.

However, though the proliferation of satellites and social media has led to a massive amount of new data for human rights investigators to parse, there’s still little labeled data to train neural networks. Looking at a satellite image of a smoke plume, “I know it’s a crime,” Lyons said. “But how do I tell the computer it’s a crime?”
satelliteimagery  myanmar  smoke  fire  neuralnetworks  deeplearning  machinelearning  nvidia 
april 2019 by fcoel

Copy this bookmark:





to read