Re: P.S. to John!!!

[P3D Jim Crowell <crowell@xxxxxxxxxxxxxxx>]

At 11:59 PM 10/15/96, P3D John W Roberts wrote:
>
>Since depth of field is dependent partly on the degree of dilation of the
>pupil, the "scale" will vary based on the amount of lighting. It would be
>interesting to find out whether people's depth judgement (one eye, head
>motionless) is better in dimmer light than in brighter light, and whether in
>estimating an absolute distance to an object they properly compensate for
>effects related to the degree of pupil dilation. (And also whether there are
>different results if they've just had an eye exam and been given those eye
>drops that force the pupils to dilate.)
>

Interesting observation.  It would be very hard to implement in human
vision because accomodation is a very weak depth cue, the pupil diameter
changes so rapidly, and the width of the blur circle is a non-monotonic
function of pupil size (if I remember right, it's smallest at around 2-3mm
& increases for larger and smaller pupil sizes).

I do know of some people at U.T. Austin who have developed a computer
vision system that computes a depth map using this principle; they take a
couple of photographs using two different (known) focal lengths and compute
the depth from the difference in blur between the two pictures at each
point.

-Jim C.

----------------------
Jim Crowell
Division of Biology
216-76
Caltech
Pasadena, CA
(818) 395-8337
jim@xxxxxxxxxxxxxxx



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