碩士 / 國立臺北科技大學 / 電資學院外國學生專班 / 105 / Unmanned Aerial Vehicle (UAV) technology has become an effective alternative
to satellite remote sensing for achieving major research breakthroughs. In this
research, methods are presented for scene detection and vehicle recognition system
using high-resolution imagery acquired through UAV platform aided with landmark
detection and recognition system to facilitate the UAV’s autonomous navigation. The
proposed systems comprises of a UAV platform that facilitates efficient autonomous
flights and captures images and real-time video streaming of ground objects using a
14-megapixel mounted camera sensor. The landmark detection and recognition system
has proved to be more robust and reliable due to landmarks color and shape properties.
ii
The UAV images acquired are preprocessed and used for scene detection to measure
the area of damaged grass fields. The UAV platform is also used as an intelligent
transportation system (ITS) for vehicle recognition which consists of two parts: a car
recognition system to automatically detect vacant packing spaces in order to facilitate
smart parking and an automatic license plate recognition (ALPR) system for vehicles
in target areas. The histogram of oriented gradients (HOG) and linear support vector
machine (SVM) methods are employed for both landmark and car recognition systems.
Subsequently, an unsupervised Iterative Self-Organizing Data Analysis Technique
(ISODATA) classification algorithm is applied to detect and differentiate
environmental classes (scene detection and interpretation) in the target or investigated
area by using the high-resolution images. This allows us to clearly differentiate the
damaged grass field and measure its area efficiently. Also, the deep convolution neural
network (CNN) architecture is used for ALPR system. Finally, the results of the
proposed methods are evaluated by comparing their results with the state-of-the-art
models.
| Identifer | oai:union.ndltd.org:TW/105TIT05706002 |
| Date | January 2017 |
| Creators | Lucky Nhlanhla Sithole, Lucky Nhlanhla Sithole |
| Contributors | Yo-Ping Huang, 黃有評 |
| Source Sets | National Digital Library of Theses and Dissertations in Taiwan |
| Language | en_US |
| Detected Language | English |
| Type | 學位論文 ; thesis |
| Format | 73 |
Page generated in 0.0272 seconds