國立虎尾科技大學 |

A novel deep learning method for object detection in multispectral images via pixel rearrangement.

紀錄類型:	書目-語言資料,手稿 : Monograph/item
正題名/作者:	A novel deep learning method for object detection in multispectral images via pixel rearrangement./
作者:	Kasi Vishwanathan, Anusha.
面頁冊數:	1 online resource (89 pages)
附註:	Source: Masters Abstracts International, Volume: 57-05.
Contained By:	Masters Abstracts International57-05(E).
標題:	Computer engineering. -
電子資源:	click for full text (PQDT)
ISBN:	9780438006454

A novel deep learning method for object detection in multispectral images via pixel rearrangement.
Kasi Vishwanathan, Anusha.

A novel deep learning method for object detection in multispectral images via pixel rearrangement. - 1 online resource (89 pages)

Source: Masters Abstracts International, Volume: 57-05.

Thesis (M.S.E.)--University of Massachusetts Lowell, 2018.

Includes bibliographical references

In this work, we study the problem of object detection in multispectral images. We present a generalized "Fully" Convolutional Neural Network (FCNN)-based deep learning system with a novel Pixel Rearrangement Technique, with reduced complexity and improved accuracy than its state-of-the-art counterparts. In particular. we (a) define a key strategy based on spectral signatures to select a set of highly informative multispectral bands for the system; (b) present the design methodology to choose the most optimized layers for the proposed FCNN architecture; (c) for the first time. introduce a pixel rearrangement technique that efficiently utilizes pixels from the network's feature maps that results into accurate pixelwise prediction images; (d) propose dual stage global and adaptive thresholding methodologies that transform the pixelwise prediction images to binary. We evaluate the proposed system for automatic building detection using the SpaceNet dataset. We use three NVIDIA GeForce GTX 1060 GPUs. Tensorflow deep learning framework to implement the proposed system. Our findings show an improvement in the performance by 0.3% in comparison to the top winning submission of SpaceNet Building Challenge II, that took place in April 2017. with an additional 43% reduction in the number of FCNN layers. Finally, we also present a comparison chart with various existing approaches to highlight our reduced computational complexity system.

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

Mode of access: World Wide Web

ISBN: 9780438006454Subjects--Topical Terms:

569006
Computer engineering.
Index Terms--Genre/Form:

554714
Electronic books.

A novel deep learning method for object detection in multispectral images via pixel rearrangement.
LDR:02607ntm a2200313Ki 4500 001 916930
005 20180928111503.5
006 m o u
007 cr mn||||a|a||
008 190606s2018 xx obm 000 0 eng d
020 $a 9780438006454
035 $a (MiAaPQ)AAI10836934
035 $a AAI10836934
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Kasi Vishwanathan, Anusha. $3 1190803
245 1 2 $a A novel deep learning method for object detection in multispectral images via pixel rearrangement.
264 0 $c 2018
300 $a 1 online resource (89 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: 57-05.
502 $a Thesis (M.S.E.)--University of Massachusetts Lowell, 2018.
504 $a Includes bibliographical references
520 $a In this work, we study the problem of object detection in multispectral images. We present a generalized "Fully" Convolutional Neural Network (FCNN)-based deep learning system with a novel Pixel Rearrangement Technique, with reduced complexity and improved accuracy than its state-of-the-art counterparts. In particular. we (a) define a key strategy based on spectral signatures to select a set of highly informative multispectral bands for the system; (b) present the design methodology to choose the most optimized layers for the proposed FCNN architecture; (c) for the first time. introduce a pixel rearrangement technique that efficiently utilizes pixels from the network's feature maps that results into accurate pixelwise prediction images; (d) propose dual stage global and adaptive thresholding methodologies that transform the pixelwise prediction images to binary. We evaluate the proposed system for automatic building detection using the SpaceNet dataset. We use three NVIDIA GeForce GTX 1060 GPUs. Tensorflow deep learning framework to implement the proposed system. Our findings show an improvement in the performance by 0.3% in comparison to the top winning submission of SpaceNet Building Challenge II, that took place in April 2017. with an additional 43% reduction in the number of FCNN layers. Finally, we also present a comparison chart with various existing approaches to highlight our reduced computational complexity system.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2018
538 $a Mode of access: World Wide Web
650 4 $a Computer engineering. $3 569006
650 4 $a Artificial intelligence. $3 559380
655 7 $a Electronic books. $2 local $3 554714
690 $a 0464
690 $a 0800
710 2 $a ProQuest Information and Learning Co. $3 1178819
710 2 $a University of Massachusetts Lowell. $3 1179049
773 0 $t Masters Abstracts International $g 57-05(E).
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=10836934 $z click for full text (PQDT)