P. M. Hillman

Alpha channel estimation in high resolution images and image sequences

For motion picture special effects, it is often necessary to take a source image of an actor, segment the actor from the unwanted background, and then composite over a new background. The standard approach requires the unwanted background to be a blue screen. While this technique is capable of handling areas where the foreground blends into the background, the physical requirements present many practical problems. This paper presents an algorithm that requires minimal human interaction to segment motion picture resolution images and image sequences. We show that it can be used not only to segment badly lit or noisy blue screen images, but also to segment actors where the background is more varied.

Issues in adapting research algorithms to stereoscopic visual effects

Many published machine vision algorithms are designed to be real-time and fully automatic with low computational complexity. These attributes are essential for applications such as stereo robotic vision. Motion Picture Digital Visual Effect facilities, however, have massive computation resources available and can afford human interaction to initialise algorithms and to guide them towards a good solution. On the other hand, motion pictures have significantly higher accuracy requirements and other unique challenges. Not all machine vision algorithms can readily be adapted to this environment. In this paper we outline the requirements of visual effects and indicate several challenges involved in using image processing and machine vision algorithms for stereo motion picture visual effects.

Sky Browser: Search for HDR Sky Maps

In a visual effects studio for movie production, sky maps play an important role for acting as a sky backdrop to a scene. The backdrop to a scene is often represented using a high-resolution sky map. This motivates the need for a large collection of sky maps to match various moods and lighting conditions. A comprehensive collection of images is not useful however, without a method of searching for desired images within that database. In this paper we define a feature space that supports an interactive search function for HDR sky maps, allowing users to find ideal images based on its appearance. The set of features are automatically extracted from the sky maps in an offline pre-processing step, and are queried in real time for progressive browsing. The system uses unsupervised learning techniques, discarding the need for labelling a large set of existing sky maps.

Hybrid Facial Model Fitting using Active Appearance Models and Contour-Based Facial Feature Location

Model-based coding techniques require a very accurate initial fit of a facial model to an image sequence in order to achieve realistic results, particularly around the eyes and mouth. This paper presents a facial model-fitting technique which uses a combination of texture analysis using an active appearance model and feature position location to give accurate initial fits. Results, based on the XM2VTS database, show this hybrid technique superior to using active appearance models alone especially in the ability to locate the eyes and mouth.