Computer-based simulations of emergencies increasingly adopt 3D graphics to visualize results and thus generate complex dynamic 3D scenes with many potentially parallel events that a↵ect large groups of virtual characters. To understand the portrayed scenario, a viewer could interactively control a flying camera or switch among a set of virtual cameras that have been previously placed at modeling time. The first solution imposes a cognitive load on the viewer that can distract him/her from the analysis task, and (s)he might miss events while moving the camera. The second solution requires additional work in the modeling phase, and even a very large number of cameras could fail to correctly frame events because of dynamic occlusions. More sophisticated automatic camera control methods could help, but the methods in the literature are designed for sequential dialogue-like events that involve at most two or three characters and therefore would not work. In this paper, we present a fully automated, real-time system that is able to monitor events in emergency simulations, select relevant events based on user-provided filtering rules, and control a virtual camera such that the events of interest are properly presented to the viewer. To illustrate how the system works in practice, we also describe the first application of automatic camera control to the domain of aviation safety.
Automatic camera control meets emergency simulations: An application to aviation safety
RANON, Roberto;CHITTARO, Luca;BUTTUSSI, Fabio
2015-01-01
Abstract
Computer-based simulations of emergencies increasingly adopt 3D graphics to visualize results and thus generate complex dynamic 3D scenes with many potentially parallel events that a↵ect large groups of virtual characters. To understand the portrayed scenario, a viewer could interactively control a flying camera or switch among a set of virtual cameras that have been previously placed at modeling time. The first solution imposes a cognitive load on the viewer that can distract him/her from the analysis task, and (s)he might miss events while moving the camera. The second solution requires additional work in the modeling phase, and even a very large number of cameras could fail to correctly frame events because of dynamic occlusions. More sophisticated automatic camera control methods could help, but the methods in the literature are designed for sequential dialogue-like events that involve at most two or three characters and therefore would not work. In this paper, we present a fully automated, real-time system that is able to monitor events in emergency simulations, select relevant events based on user-provided filtering rules, and control a virtual camera such that the events of interest are properly presented to the viewer. To illustrate how the system works in practice, we also describe the first application of automatic camera control to the domain of aviation safety.File | Dimensione | Formato | |
---|---|---|---|
CG-Elevaid-corrected.pdf
accesso aperto
Descrizione: Articolo principale
Tipologia:
Documento in Post-print
Licenza:
Creative commons
Dimensione
12.84 MB
Formato
Adobe PDF
|
12.84 MB | Adobe PDF | Visualizza/Apri |
1-s2.0-S0097849315000291-main.pdf
non disponibili
Tipologia:
Versione Editoriale (PDF)
Licenza:
Non pubblico
Dimensione
7.65 MB
Formato
Adobe PDF
|
7.65 MB | Adobe PDF | Visualizza/Apri Richiedi una copia |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.