A fundamental goal of computer vision research is the development of systems capable of carrying out scene interpretation using knowledge of the expected scene. Here, we describe PSEIKI, a framework for expectation-driven interpretation of image data. PSEIKI performs expectation-driven processing by matching elements, such as edges and regions, detected in an image with model-elements front a supplied expected scene. PSEIRI builds abstraction hierarchies in image data using cues taken from the supplied abstractions in the expected scene. Hypothesized abstractions in the image data are geometrically compared with the known abstractions in the expected scene; the metrics used for these comparisons translate into belief values. The Dempster-Shafer formalism is used to accumulate beliefs for the synthesized abstractions in the image data. For accumulating belief values, a computationally efficient variation of Dempster’s rule of combination is developed to enable the system to deal with the overwhelming amount of information present in most images. This variation of Dempster’s rule allows the reasoning process to be embedded into the abstraction hierarchy by allowing for the propagation of belief values between elements at different levels of abstraction. PSEIKI has been implemented as a 2-panel, 5-level blackboard in OPS83. The operation and implementation of the blackboard’s knowledge sources are described in detail. Control aspects of the blackboard’s scheduler and distributed monitor are also described. Finally, an experiment in which PSEIKI was used to aid in the navigation of an autonomous mobile robot will be described. PSEIKI was used to provide sensory feedback to update the estimates of the robot’s position and orientation as if traveled in a known environment.
Date of this Version