Computer vision systems (S. Lu)

In recent years, computer vision research has attempted to develop a machine or robot intelligence that can sense its environment, understand or interpret what it senses and recognize the objects in a digitized image for decision making or action.

Our research focuses on (a) time-efficient early processing algorithms, (b) determining the orientation and position of 3-D objects relative to the machine coordinate system or the real world coordinate system; (c) from a digitized image (intensity data, range data) whether or not an object in the image corresponds to a distinguishable object whose symbolic representation is stored in a database, and (d) for objects unrecognizable or unknown to the system, to synthesize the information from the acquired images and to store the processed information in a form recognizable by the system in the future.

Revised: 94.10.25