Cortical Responses to Illusory and Surface Color

Loading...
Thumbnail Image

Authors

Coia, Andrew J.

Issue Date

2016

Type

Dissertation

Language

Keywords

color , hdEEG , illusion , source localaization , vision , watercolor effect

Research Projects

Organizational Units

Journal Issue

Alternative Title

Abstract

In order to see the world as we do our visual system constantly performs many amazingfeats that are largely unknown to us, the viewers. In order to create and update a seamless andcoherent representation of the world, our eyes and brain have to deal with many obstacles thatcould interfere with object recognition such as occluding blood vessels in our eyes, layers ofoccluding retinal neurons, the blind spot that is devoid of photoreceptors, objects and shadowsoccluding other objects, and retinal inhomogeneities. Ultimately, we obtain information thatencodes relative brightness and colors that allow us to recognize different objects and surfaces.How is it that we obtain the information to assign color to object surfaces? It has beenhypothesized that we predominantly extract information about the color of an object from thespectral contrast at the edges and fill in the remaining areas (if the edge information isconsistent with that from a uniform surface). These ideas have been put forth to explain strikingvisual illusions such as the Cornsweet Illusion, neon color spreading, and the watercolor effect.All of these illusions exist in both color and brightness domains. Thus, even though it may notbe intuitively obvious, edges play a key role in our final percept of what color we see whenviewing an object. It is said that we ‘perceptually fill in’ color from edges similarly to how ourvisual system fills in the blind spot in our eye. Of course it should be noted that spectralinformation may also be available from within the regions away from the edges and thisinformation may also play an important role in the perception of surface color. The theory thatedge information alone determines color and that internal surface information is discardedimplies that there is no difference in neural computation of physical surface color and edgeiinduced color spreading. Initial studies of this question suggest that illusory colors and actualsurface colors may in fact show important differences. One goal of the present study is toemploy both psychophysics and electrophysiology to determine under what conditions illusorycolors and surface colors are differentiated to provide insights into the fundamental processesinvolved in surface color perception.If filling-in is indeed important for surface color perception, then how might color fillinginoccur? One possibility is that information from the edges might be propagated in a feedforwardmanner that passively spreads until another edge or contradicting information isencountered. Another possibility is that edge information could be relayed to higher order formcenters and then surface colors reconstructed from feedback from those higher regions. Thequestion of feed-forward vs. feedback mechanisms of perception has been debated by scientistssince at least the time of Hering and Helmholtz, Hering favoring a bottom up interpretationwhile Helmholtz argued that our visual system performs ‘unconscious inferences.’ It is currentlyappreciated that a relatively sparse sampling of visual information from the world results in arich visual percept, supporting the idea that much of our vision is actually reconstructed frompast experience. Another goal of the present study is to apply electrophysiological methods(hdEEG) to determine the relative importance of feed-forward and feedback mechanisms insurface color perception. This work expands on previous research which developed a method ofmeasuring the watercolor illusion with single channel VEP.This dissertation provides a literature review of the background and significance of theproblems, presents preliminary data, outlines the series of proposed experiments, and lastly theresults of the proposed experiments.

Description

Citation

Publisher

License

In Copyright(All Rights Reserved)

Journal

Volume

Issue

PubMed ID

DOI

ISSN

EISSN