CAPSLOCKSTUCK
Spaced Out Lunar Tick
- Joined
- Feb 26, 2013
- Messages
- 8,578 (2.12/day)
- Location
- llaregguB...WALES
System Name | Party On |
---|---|
Processor | Xeon w 3520 |
Motherboard | DFI Lanparty |
Cooling | Big tower thing |
Memory | 6 gb Ballistix Tracer |
Video Card(s) | HD 7970 |
Case | a plank of wood |
Audio Device(s) | seperate amp and 6 big speakers |
Power Supply | Corsair |
Mouse | cheap |
Keyboard | under going restoration |
A piece of computer software has been developed that was able to uncover almost 7,000 previously undiscovered craters in a matter of hours.
The finding was made by a team of researchers led by Ari Silburt at Penn State University and Mohamad Ali-Dib at the University of Toronto.
They fed 90,000 images of the moon's surface into an artificial neural network (ANN).
ANNs try to simulate the way the brain works in order to learn and can be trained to recognise patterns in information.
The team's ANN had been taught to categorise images and identify craters larger than five kilometres (three miles) in diameter.
When the machine was tested it located 6,883 new craters, doubling the total number of known craters of this size.
Researcher believe that, with further training, the system could be used to accelerate crater counting even more.
The algorithm was trained to identify the edges of a crater, which it then checked against a database of previously discovered craters.
It used this information to confirm that the shape of the new object conformed to known crater shapes.
This allowed it to distinguish craters from other geological shapes, including mountains and ridges.
Many of the images it used were captured by Nasa's Lunar Reconnaissance Orbiter (LRO), which was launched via an Atlas V rocket in 2009.
New Scientist
The finding was made by a team of researchers led by Ari Silburt at Penn State University and Mohamad Ali-Dib at the University of Toronto.
They fed 90,000 images of the moon's surface into an artificial neural network (ANN).
ANNs try to simulate the way the brain works in order to learn and can be trained to recognise patterns in information.
The team's ANN had been taught to categorise images and identify craters larger than five kilometres (three miles) in diameter.
When the machine was tested it located 6,883 new craters, doubling the total number of known craters of this size.
Researcher believe that, with further training, the system could be used to accelerate crater counting even more.
The algorithm was trained to identify the edges of a crater, which it then checked against a database of previously discovered craters.
It used this information to confirm that the shape of the new object conformed to known crater shapes.
This allowed it to distinguish craters from other geological shapes, including mountains and ridges.
Many of the images it used were captured by Nasa's Lunar Reconnaissance Orbiter (LRO), which was launched via an Atlas V rocket in 2009.
New Scientist