碩士 / 國立雲林科技大學 / 電機工程系 / 104 / “Location Awareness” and “Destination Navigation” are fundamental to versatile ubiquitous wearable device’s applications and service, this paper develops an infrastructureless indoor Augmented Reality Navigation (ARN) wearable device. It can lay 3D virtual navigation directions over what users are actually seeing in front of themselves in real world without deploying any infrastructure or markers. In order to make it more precise, reliable, and instantaneous, this work proposes “optical-flow-scene indoor positioning” based on dead reckoning to improve “feature-optical-flow indoor positioning” and “optical-flow-field indoor positioning”, proposes “adjacent-list-coordinate path planning” based on Dijkstra algorithm to improve “adjacent-matrix-node path planning” and “adjacent-list-node path planning”, proposes “wall-floor-boundary image registration” based on floor segmentation to improve “End-Dot-EdgeLine image registration” and “panoramic-image-segmentation image registration”, proposes “optical-flow-inertial pose estimation” based on markerless tracker to improve “specific-marker-tracking pose estimation” and “nature-feature-tracking pose estimation”, and proposes “small-angle-approximation projective transformation” based on Homography matrix to improve “Homogeneous-Homography-matrix projective transformation” and “Decomposed-Homograph-matrix projective transformation”. Implementation results show this thesis featuring these proposed 5 methods has higher accuracy and less latency than conventional well-known ARN methods. Besides, this thesis featuring these proposed 5 methods has been implemented into Android wearable device seamlessly and smoothly. This work is suitably applied to versatile outdoor and indoor wearable navigation applications, like site directions, event guidance, merchandise seeking, social searching, and so on.
| Identifer | oai:union.ndltd.org:TW/104YUNT0441024 |
| Date | January 2016 |
| Creators | WANG,PO-KAI, 王柏凱 |
| Contributors | HO,CHIAN-CHENG, 何前程 |
| Source Sets | National Digital Library of Theses and Dissertations in Taiwan |
| Language | zh-TW |
| Detected Language | English |
| Type | 學位論文 ; thesis |
| Format | 53 |
Page generated in 0.0088 seconds