Vision

Images convey multiple meanings that depend on the context in which the viewer perceptually organizes the scene. In her tenth anniversary edition of her doctoral dissertation, she examines the influence of illusory contours and figure-ground segmentation on contrast processing. She finds powerful empirical evidence demonstrating that global context can alter local contrast processing, contrary to the classical feed-forward model. She discusses experiments, utilizing rich stimuli context, that probe the details of global/local interactions. She introduces models incorporating dynamic filters as a mechanism for modeling unconscious inference. Modern computer vision techniques of image segmentation and labeling inspired by this original work are discussed in the forward. Use of fuzzy inference systems and spatial-taxons for solving machine vision problems are described in the context of the emerging field of visual-taxometrics.

Autorentext

A scientist, inventor and serial entrepreneur, Lauren Barghout holds a Ph.D. & M.S. in Vision Science from UC Berkeley and a BA degree in Physics from Hampshire College. She is an expert in the use of fuzzy set theory in representing Visual Gestalts. Her work in human & machine vision has cited by companies such as Microsoft, Hitachi and Google.