Using Occlusion Calculi to Interpret Digital Images

This paper reports on an investigation using occlusion calculi to interpret digital images. Using a minimal set of digital, region-relation detectors, and assuming a continuous interpretation of physical space, we show how existing calculi can be augmented and embedded in the Event Calculus to interpolate and recover a larger set of occlusion relations than are otherwise available at the basic detector level.1

• Type: Conference
• Year: 2006
• Where: ECAI
• Authors: David A. Randell, Mark Witkowski
• Keywords: Artificial Intelligence, Basic Detector Level.1, ECAI 2006, Occlusion Calculi, Occlusion Relations