國立虎尾科技大學 |

An Optimized Object Detection Model for Unmanned Aerial Vehicles.

紀錄類型:	書目-語言資料,手稿 : Monograph/item
正題名/作者:	An Optimized Object Detection Model for Unmanned Aerial Vehicles./
作者:	Onifade, Temilola O.
面頁冊數:	1 online resource (70 pages)
附註:	Source: Masters Abstracts International, Volume: 85-12.
Contained By:	Masters Abstracts International85-12.
標題:	Robotics. -
電子資源:	click for full text (PQDT)
ISBN:	9798383180648

An Optimized Object Detection Model for Unmanned Aerial Vehicles.
Onifade, Temilola O.

An Optimized Object Detection Model for Unmanned Aerial Vehicles. - 1 online resource (70 pages)

Source: Masters Abstracts International, Volume: 85-12.

Thesis (M.S.)--University of Louisiana at Lafayette, 2023.

Includes bibliographical references

Unmanned Aerial Vehicles (UAVs) have impacted a wide range of sectors, including surveillance, agriculture, photography, and military among many others. UAVs applications have improved with the expansion of artificial intelligence in research and development. Artificial Intelligence techniques like speech recognition, computer vision, and facial recognition have had specific improvement in UAVs. Computer Vision for obstacle avoidance in many applications including robotics, autonomous vehicles and UAVs have become common. In this study, computer vision is explored to improve object detection and collision avoidance on UAVs. The most costly and time-consuming part of distribution logistics is package delivery. UAVs have emerged as a promising solution to reduce delivery times and overall delivery costs. Detecting obstacles and avoiding collision is an important safety feature the drone requires for this delivery process.In this thesis, a new labeled dataset is used for increased accuracy compared to currently used datasets. Obstacles identified are birds, trees, drones, airplanes, and helicopters. While the drone is flying, a camera will be used to gather information for the surrounding environment, which is then evaluated using the latest version of You Only Look Once (YOLO), YOLOv8 and Faster Region-based Convolution Neural Network (R-CNN). According to literature, these are the best object detection models. YOLO has been confirmed to be the fastest algorithm and with a high accuracy for real-time applications. With great inference speed, YOLO struggles to detect smaller objects. Here, Faster R-CNN can achieve this with a lower speed tradeoff. The accuracy and speed of the models are important, and some performance metrics will be discussed to evaluate the models. If the models are not accurate, collisions could occur, which would result in damage. This project achieves comparable accuracy by the models used. This improves the overall performance of the delivery drone for faster delivery with lower cost.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2024

Mode of access: World Wide Web

ISBN: 9798383180648Subjects--Topical Terms:

561941
Robotics.
Subjects--Index Terms:

Evaluation metricsIndex Terms--Genre/Form:

554714
Electronic books.

An Optimized Object Detection Model for Unmanned Aerial Vehicles.
LDR:03435ntm a22004097 4500 001 1148296
005 20240924101843.5
006 m o d
007 cr bn ---uuuuu
008 250605s2023 xx obm 000 0 eng d
020 $a 9798383180648
035 $a (MiAaPQ)AAI30426922
035 $a AAI30426922
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Onifade, Temilola O. $3 1474234
245 1 3 $a An Optimized Object Detection Model for Unmanned Aerial Vehicles.
264 0 $c 2023
300 $a 1 online resource (70 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: 85-12.
500 $a Advisor: Bayoumi, Magdy A.
502 $a Thesis (M.S.)--University of Louisiana at Lafayette, 2023.
504 $a Includes bibliographical references
520 $a Unmanned Aerial Vehicles (UAVs) have impacted a wide range of sectors, including surveillance, agriculture, photography, and military among many others. UAVs applications have improved with the expansion of artificial intelligence in research and development. Artificial Intelligence techniques like speech recognition, computer vision, and facial recognition have had specific improvement in UAVs. Computer Vision for obstacle avoidance in many applications including robotics, autonomous vehicles and UAVs have become common. In this study, computer vision is explored to improve object detection and collision avoidance on UAVs. The most costly and time-consuming part of distribution logistics is package delivery. UAVs have emerged as a promising solution to reduce delivery times and overall delivery costs. Detecting obstacles and avoiding collision is an important safety feature the drone requires for this delivery process.In this thesis, a new labeled dataset is used for increased accuracy compared to currently used datasets. Obstacles identified are birds, trees, drones, airplanes, and helicopters. While the drone is flying, a camera will be used to gather information for the surrounding environment, which is then evaluated using the latest version of You Only Look Once (YOLO), YOLOv8 and Faster Region-based Convolution Neural Network (R-CNN). According to literature, these are the best object detection models. YOLO has been confirmed to be the fastest algorithm and with a high accuracy for real-time applications. With great inference speed, YOLO struggles to detect smaller objects. Here, Faster R-CNN can achieve this with a lower speed tradeoff. The accuracy and speed of the models are important, and some performance metrics will be discussed to evaluate the models. If the models are not accurate, collisions could occur, which would result in damage. This project achieves comparable accuracy by the models used. This improves the overall performance of the delivery drone for faster delivery with lower cost.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2024
538 $a Mode of access: World Wide Web
650 4 $a Robotics. $3 561941
650 4 $a Information science. $3 561178
650 4 $a Computer engineering. $3 569006
650 4 $a Electrical engineering. $3 596380
653 $a Evaluation metrics
653 $a Convolution Neural Network
653 $a Object detection
653 $a Unmanned Aerial Vehicles
653 $a You Only Look Once
655 7 $a Electronic books. $2 local $3 554714
690 $a 0544
690 $a 0464
690 $a 0723
690 $a 0771
710 2 $a University of Louisiana at Lafayette. $b Electrical Engineering. $3 1474235
710 2 $a ProQuest Information and Learning Co. $3 1178819
773 0 $t Masters Abstracts International $g 85-12.
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30426922 $z click for full text (PQDT)