Re: weighting depth clues

[T3D Scott Langill <slangill@xxxxxxxxx>]

On Fri, 18 Jul 1997 John Bercovitz <bercov@xxxxxxxxxxx> wrote

> I wonder how depth clues are weighted.  For my talk at NSA
.....
> (The heights of parts of the car appear unaffected by the
> stretch condition which is expected from the geometric
> reconstruction.)  So what can you tell us about how depth
> clues work together, Jim?  What overrides what under what
> conditions?  Is the internal model sometimes a stronger
> clue and sometimes a weaker clue according to how well
> we know the object?

As to the last point, the erroneous judgements made in distorted rooms of
the "Mystery Spot" variety would seem to indicate that the internal model
is fairly plastic, while the inability of people to see stereo reversed
faces as reversed in depth would indicate that there are some rudimentary
constants. The visual cliff experiments (where infant creatures avoid the
part of a surface painted with gradients making it look further away or
deeper) have shown that some depth perception mechanisms are innate (at
least in precocious species) and may be the result of a specific type of
neuron rather than a higher order mental judgement (like your car). The
resolution of depth in random patterns shown in the Bell Labs experiments
would tend to point to an innate ability as well.

It is an interesting question though. Cues to depth and distance
have been enumerated in psychophysics. Each illusion of depth, distance,
and size is presumably one or more cues outweighing one or more other
cues. But I don't think that anyone has attempted to analyze these
illusions to see if there is consistency or whether an ordinal ranking of
relative strength is possible.

I remember hearing in a lecture before about the relative weakness of
vertical disparity and cues and that it was presumed due to the fact that
the vast majority of visual information and visual judgements that humans
make have to do with the horizontal and with things within a few degrees
of the visual horizon. Perhaps with other species (like birds) this would
be different.

@#!*@#!*@#!*@#!*@#!*@#!*@#!*@#!*@#!*@#!*@#!*@#!*@#!*@#!*@#!*@#!*@#!*@#!*
Scott W. Langill                                   Arbeit mach das leben
Special Edition          slangill@xxxxxxxxx        suesse, aber faulheit
Washington, D.C.                                 staerkt die gliederung.
@#!*@#!*@#!*@#!*@#!*@#!*@#!*@#!*@#!*@#!*@#!*@#!*@#!*@#!*@#!*@#!*@#!*@#!*



------------------------------

End of TECH-3D Digest 182
*************************