<title>Active vision and image/video understanding systems for intelligent manufacturing</title>

摘要

A new generation of industrial robots needs to have reliable perceptual systems that are similar to human vision. Human vision is based on the principles of image understanding and active vision. Both principles are possible in the form of Network-Symbolic systems. An Image/Video Analysis that is based on Network-Symbolic principles significantly differs from a traditional image analysis. Instead of precise computations of 3-dimensional models, such a system converts image into an "understandable" relational format similar to knowledge models. It is hard to use geometric operations for processing of natural images. Instead, the brain builds a relational network-symbolic structure of visual scene, using different clues to set up the relational order of surfaces and objects with respect to the observer and to each other. Spatial order can be represented as a connection graph. There is a generic logic of 3-D structures, which is based on relational changes of object views in the visual or object buffers. In Network-Symbolic Models, the derived structure and not the primary view is a subject for recognition. Such recognition is not affected by local changes and appearances of the object as seen from a set of similar views. Integrated into the industrial robotic systems, Network-Symbolic models can intelligently interpret images and video.