Solving the Brightness-from-Luminance Problem: A Neural Architecture for Invariant Brightness Perception

摘要

The spatial distribution of light that constitutes the input to our eyes is the foundation of all visual functions, such as perception of brightness, color, texture, form, and 3-D organization. The perception of brightness may perhaps appear to be the simplest of all functions: The most natural initial explanation of why surface A appears brighter than surface B is that more light arrives into our eyes from surface A than from B. However, as we will show in the following, the relation of luminance (which is a physical variable involving the amount of light energy arriving at the retina) and brightness (which is a psychological variable denoting perceived intensity of light) is much more complicated. The brightness-from luminance problem is the following: find the mapping that transforms any given spatial distribution of luminance into the corresponding spatial distribution of brightness. The problem is generally solved for the simple visual situation involving a bright patch on a dark background. Increasing the luminance of the patch causes it to look increasingly brighter, but in a nonlinear manner. In more complicated visual situations containing several surfaces, their brightnesses may be predicted by taking logarithms, or power functions, of their luminance. To summarize, there are at least two factors that make the relation of brightness and luminance a problem: Illumination discounting and contextual dependence. We will present a neural network architecture that deals with both issues. Keywords: Mathematical models. (KT/AW)