A Constraint-Based Autonomous 3D Camera System

Abstract

Camera control techniques for interactive digital entertainment (IDE) are reaching their limits in terms of capabilities. To enable future growth, new methods must be derived to address these new challenges. Existing academic research into camera control is typically devoted to cinematography and guided exploration tasks, and is not directly applicable to IDE. This paper describes a novel application of constraint satisfaction in the design of a camera system that addresses the unique and difficult challenges of IDE. It demonstrates a specialized constraint solver that exploits the spatial structure of the problem, enabling the real-time use of the camera system. The merit of our solution is highlighted by demonstrating the computational efficiency and ability to extend the cameras capabilities in a simple and effective manner.

Automatic Camera Control: A Dynamic Multi-Objective Perspective

“…By definition, none of the contained solutions can dominate any other. 4 For a given test point and a given front, we find all members of the reference front that dominate it or are dominated (note that a point can either dom-inate or be dominated by one or multiple points of the reference front, but not both). Points that are incomparable (neither dominated nor dominate) are assigned a value of 0.…”

“…The different objective functions are combined linearly to produce a single objective function which can be optimised either in a static environment or in a dynamic one. A variety of algorithms have been employed in the two cases including, among others, Genetic Algorithms [15,17] and Particle Swarm Optimisation [6] for static scenes, and Hill Climbing [4] and Artificial Potential Fields [7] for real-time optimisation in dynamic scenes. The first two approaches are used to generate still images with specific composition characteristics, while the last two are designed to animate a camera in real-time interactive virtual environment -e.g.…”

“…the game is played. These approaches have been successfully employed in games [4,5] and are particularly suitable for real-time tracking of optimal cameras in interactive and unpredictable virtual environments. However, due to their reactive nature, they are ill suited in situations where the optimum at a certain moment is not necessary the best solution.…”

Combining local and global optimisation for virtual camera control

“…Beckhaus et al [4] first introduced local search algorithms to camera control; their system employs Artificial Potential Fields to generate collision-free camera paths through a virtual environment. Bourne and Sattar [7] proposed a system that employs sliding octrees to guide the camera to the optimal camera configuration. Burelli and Jhala [9] extended these two approaches to include frame composition and support multiple-object tracking.…”

“…A common technique to detect occlusion consists of casting a series of rays between the object of interest and the camera [8], [7]. Marchand and Courty [22] generate a bounding volume containing both the camera and the object of interest and check whether other objects intersect this volume.…”