政大機構典藏-National Chengchi University Institutional Repository(NCCUR):Item 140.119/149450

政大典藏 > College of Informatics > Department of Computer Science > Periodical Articles > Item 140.119/149450

Please use this identifier to cite or link to this item: https://nccur.lib.nccu.edu.tw/handle/140.119/149450

Title:	Multi-view approach for drone light show
Authors:	紀明德 Chi, Ming-Te;Weng, Kai-Chun;Lin, Shu-Ting;Hu, Chen-Chi;Soong, Ru-Tai
Contributors:	資訊系
Keywords:	3D ambigram;Drone light show;Integer programming
Date:	2023-11
Issue Date:	2024-01-29 09:45:26 (UTC+8)
Abstract:	Drones, or precisely quadrotors, have been increasingly used in the field of robotics and also in entertainment. Coordinated multiple drones that form visual presentations through their equipped LEDs are known as drone light shows (Waibel, M., Keays, B., Augugliaro, F.: in Drone shows: Creative potential and best practices. ETH Zurich, 2017). Such a performance offers visual enjoyment for a large audience, particularly in festivals. However, the majority of current drone light shows are manually coordinated by personnel using software. Drone light shows also have limited viewing range, thereby preventing the audience from getting a good view of the actual performance. This study proposes a method to provide multiple visual presentations in accordance with multiple viewing angles. We use visual hull to filter out the candidate areas that form input images, and takes projection error and classification values as weight for optimization. Consequently, the proposed method reduces the number of drones needed to form a multi-view structure for visual presentations. Furthermore, to meet the demands of performing the animation in multi-view structure, we implement the flight algorithm to locate the most suitable corresponding points between two different structures, and then generate the shortest flight paths without collision. Experiments conducted in our simulator provide additional insights and discussions, and each factor is visualized to provide an improved understanding of our approach for multi-view drone light shows.
Relation:	The Visual Computer, Vol.39, pp.5797-5808
Data Type:	article
DOI link:	https://doi.org/10.1007/s00371-022-02696-8
DOI:	10.1007/s00371-022-02696-8
Appears in Collections:	[Department of Computer Science ] Periodical Articles

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	93	View/Open

社群 sharing

Loading...