國立虎尾科技大學 |

Investigation of Deep Neural Network Image Processing for Cubesat Size Satellites.

紀錄類型:	書目-語言資料,手稿 : Monograph/item
正題名/作者:	Investigation of Deep Neural Network Image Processing for Cubesat Size Satellites./
作者:	Braun, Adam D.
面頁冊數:	1 online resource (73 pages)
附註:	Source: Masters Abstracts International, Volume: 57-06.
Contained By:	Masters Abstracts International57-06(E).
標題:	Engineering. -
電子資源:	click for full text (PQDT)
ISBN:	9780438107182

Investigation of Deep Neural Network Image Processing for Cubesat Size Satellites.
Braun, Adam D.

Investigation of Deep Neural Network Image Processing for Cubesat Size Satellites. - 1 online resource (73 pages)

Source: Masters Abstracts International, Volume: 57-06.

Thesis (M.S.)--Morehead State University, 2018.

Includes bibliographical references

Cubesats first became effective space-based platforms when commercial-off-the-shelf hardware became cheap, powerful, and small enough for groups with low budgets to perform truly useful missions in space. With the growing use of embedded systems for consumers, billions of dollars being poured into artificial intelligence research, and the production of commodity hardware capable of utilizing this AI technology in consumer products, small form factor processors are now available that can multiply the computational capabilities of current Cubesat designs. Some of these embedded processors, such as the Nvidia Jetson TX2 and the Movidius Neural Compute Stick, have been specifically developed to run deep learning algorithms for Earth-based embedded systems. Since Cubesats tend to follow technology capabilities and trends of Earth-based systems, the current technology on the commercial market now allows even simple CubeSats to utilize AI algorithms. These computationally intensive AI algorithms, such as Deep Learning, are now possible to use in power limited devices with these off-the-shelf embedded systems with GPU integrations or special chip architectures for Deep Neural Network computations. This project investigates some uses of Deep Neural Networks on Cubesats and the capabilities of some low-cost, off-the-shelf hardware that could be implemented on a Cubesat to do the computations for Deep Neural Networks. An image inferencing system is developed and benchmarked on hardware that is small, lightweight, and low enough power to be used on a Cubesat and it is determined that Deep Neural Networks can be practically used on small satellites in some cases.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2018

Mode of access: World Wide Web

ISBN: 9780438107182Subjects--Topical Terms:

561152
Engineering.
Index Terms--Genre/Form:

554714
Electronic books.

Investigation of Deep Neural Network Image Processing for Cubesat Size Satellites.
LDR:02913ntm a2200349Ki 4500 001 916891
005 20180928111502.5
006 m o u
007 cr mn||||a|a||
008 190606s2018 xx obm 000 0 eng d
020 $a 9780438107182
035 $a (MiAaPQ)AAI10790236
035 $a (MiAaPQ)morehead:10177
035 $a AAI10790236
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Braun, Adam D. $3 1190754
245 1 0 $a Investigation of Deep Neural Network Image Processing for Cubesat Size Satellites.
264 0 $c 2018
300 $a 1 online resource (73 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Masters Abstracts International, Volume: 57-06.
500 $a Adviser: Jeffrey A. Kruth.
502 $a Thesis (M.S.)--Morehead State University, 2018.
504 $a Includes bibliographical references
520 $a Cubesats first became effective space-based platforms when commercial-off-the-shelf hardware became cheap, powerful, and small enough for groups with low budgets to perform truly useful missions in space. With the growing use of embedded systems for consumers, billions of dollars being poured into artificial intelligence research, and the production of commodity hardware capable of utilizing this AI technology in consumer products, small form factor processors are now available that can multiply the computational capabilities of current Cubesat designs. Some of these embedded processors, such as the Nvidia Jetson TX2 and the Movidius Neural Compute Stick, have been specifically developed to run deep learning algorithms for Earth-based embedded systems. Since Cubesats tend to follow technology capabilities and trends of Earth-based systems, the current technology on the commercial market now allows even simple CubeSats to utilize AI algorithms. These computationally intensive AI algorithms, such as Deep Learning, are now possible to use in power limited devices with these off-the-shelf embedded systems with GPU integrations or special chip architectures for Deep Neural Network computations. This project investigates some uses of Deep Neural Networks on Cubesats and the capabilities of some low-cost, off-the-shelf hardware that could be implemented on a Cubesat to do the computations for Deep Neural Networks. An image inferencing system is developed and benchmarked on hardware that is small, lightweight, and low enough power to be used on a Cubesat and it is determined that Deep Neural Networks can be practically used on small satellites in some cases.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2018
538 $a Mode of access: World Wide Web
650 4 $a Engineering. $3 561152
650 4 $a Aerospace engineering. $3 686400
650 4 $a Artificial intelligence. $3 559380
655 7 $a Electronic books. $2 local $3 554714
690 $a 0537
690 $a 0538
690 $a 0800
710 2 $a ProQuest Information and Learning Co. $3 1178819
710 2 $a Morehead State University. $b Science and Technology. $3 1190755
773 0 $t Masters Abstracts International $g 57-06(E).
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=10790236 $z click for full text (PQDT)