P-2B-6 両眼立体視における重畳刺激の奥行き定位(2001年度 日本基礎心理学会第20回大会優秀発表賞)

The Japanese Psychonomic Society

NII-Electronic Library Service

TheJapanesePsychonomicSociety

1vee

1at)anese

Jb"'"nqi

Qt'

FVy'chono"iic

2002,

VoL21,

No.

I,S5 56

Science

Prizewinner's

_{SummaryP-2B-6}

Depth

_perception

of

spatio

superimposed

_gratings

:

-temporal

_{configuration}

'

Junji

YANAGi

Chiba

Uiniversity'

The

effects

of

This

study

concerned

the

effects

of

spatio･temporal

configuration

ef

superimposed

_grating

stimuli on

the

depth

attractionfrepulsion

phenomena.

When

a sinusoidal

grating

of ambiguous

depth

(a

"targeti')

was superimposed

on

another

grating

ef

unambiguous

depth

<an

"inducer"),

_the

inducer

had

either an attractive or repulsive effect on

the

target's

_depth

_perception,

depending

on

the

spatial-frequency

relationship.

In

the

first

experiment

depth

attraction

still

occurred

but

depth

repulsion vanished when a

target

did

not

temporally

overlap with an

inducer.

The

result suggested

that

different

mechanisms contributed

the

attraction and repulsion.

In

the

second experiment a

target

of

the

same

height

as an

inducer

was always attracted, regardless

of

the

other

_parameters.

The

result

suggested

that

there

was a contribution of shape

processing.

Key

words :

binocular

stereopsis,

_grating,

depth

attraction,

depth

repulsion

Introduction

When

a sinusoidal

_grating

with an arnbiguous

depth

(a

"target")

_js

superimposed on another with an unambiguously

_perceived

depth

(an

"inducer"),

the

depth

of

the

target

can

be

perceived

as

either

in

the

same

direction

as

the

inducer

("attraction"),

or

in

the

opposite

direction

<"repulston"),

depending

un

the

spatial-frequency relationship of

the

_gratings

(Yanagi

&

Sato,

1999),

A

depth

attraction occurred

when

_the

spatial

frequency

of

the

inducer

was

lower

than

that

of

the

target.

A

depth

repulsion

occurred

when

the

spatial

frequency

of

_the

inducer

was

higher

than

that

of

the

target.

Smallman

(1995)

reported

that

a

higher-frequency

inducer

had

an attractive

effect

on a

lower-frequency

target.

But

the

spatio-temporal

configuration of

the

stimu!us used

by

Small-man

(1995)

was

different

from

that

used

in

our study

(Yanagi

&

Sato,

1999).

The

_present

study

investigat-ed

the

infiuence

of

the

spatio-temporal

configuration

on

the

depth

attraction/repulsion.

*

_Department

_of

_Psychelogy,

_Division

of

Behavior-al

Scienees,

Faeulty

of

Letters,

Chiba

University.

1-33

Yayoi-cho,

Inage-ku,

Chiba,

263-8522

Methods

An

inducer

(either

O.23

or

1.67

cldeg) and a

target

(either

O.93

or

O.42c!deg)

were

_presentecl

on a

gamma-corrected

munitor

temporally

interleaved

at

a

refresh

rate

of

100Hz.

Each

grating

randonily

changcd

its

spatial

phase

at evcry

frame.

Thc

binocu-lar

_phase

disparity

of a

target

was approximately

18e

deg

and

the

inducer

had

an either crossed or uncros･ sed

_phase

disparity,

The

subjects were required

to

judge

the

depth

direction

("Near"

or

"Far")

of

the

target

with respect

to

the

fixat'ion

_pojnt.

Results

The

response raties of

the

"Near"

judgments

were

plotted

as a

function

of a

target's

_phase

disparity

according

to

the

inducer's

depth

conditions

and

each resulting

_plot

was

fitted

with a

logistic

function.

A

target's

phase

disparity

where

the

""Near" response

ratio

became

50%

was

determined

on

the

fitted

logis-tic

function

of

the

crossed-disparity

inducer

and subtracted

from

that

of

the

uncrossed-disparity

in-ducer.

This

value can

be

regarded as an

indicator

of

the

induction

effect

of

the

inducer,

i.e.,

the

"depth

induction

index".

A

positive

or

negative

value

oi

this

The Japanese Psychonomic Society

NII-Electronic Library Service

TheJapanesePsychonomic Society

56

The

Japanese

Journalof

PsychonomicScience

Vol.

21,

No.

1

(a)

X

?

E

g

e

E

-a

e

n

14012010080604020

o-20

o

TargetlLnducerSF(eldeg}

-o-

ti,g3io.23

-･

i- O,4211.67

E."21.]li;;lr.=T=2'"

'i:;;:';';vT

(b}

ts

50

2

4o

=

30

.9

6

20

g

lo

so

=

'a

-10

o

a

-2o

iX

-TargetSF=O.42ddeg [nducerSF･=1.67eldeg TargetHeigNfeOarcmLn

×

-opa

Figure

1.

{a)

The

depth

spatial-frequency

combinations.

100

200

300

l

+4mu

SOA

(msec}

inductien

index

as

a

function

of

<b)

The

depth

induction

60

70

.80

90

Tmp Tmp Tmp

'Tmp

lnducer

Height

(arc

min}

SOA.

Different

symbols

corresponcl

to

different

index

as a

function

of

inducer

height,

index

indicates

the

occurrence

of

depth

attraction

or

repulsion

respectively.

Figure

1(a}

il]ustrates

the

effect of stimulus onset asynchrony

(SOA)

between

target

and

inducer.

Both

the

target

and

inducer

were

_presented

for

200msec.

In

a

typical

conditien

for

depth

attraction

(open

circles),

depth

attraction

occurred

when

two

gratings

were nbt

temporally

superimposed

(SOA=200

msec)

.

The

attraction was

preserved

even when a

brief

blank

was

inserted

between

the

twe

_gratings

(SOA=300

msec),

In

contrast

in

a

typical

condition

for

depth

repulsion

(filled

_triangles),

_the

repulsion

vanished

when

the

two

_gratings

were not

temporally

super-imposed.

In

acldition,

depth

attraction occurred when

a

brief

blank

was

inserted.

Figure1(b)

shows

how

in

a

typieal

conditien

for

depth

repulsion

the

inducer

height

affects

the

depth

induction

index.

Depth

attraction occurred

when

the

inducer

had

the

same

height

as

the

target.

The

index

changed

from

_positive

(attraction)

to

negative

(repulsion),

as

the

inducer

height

increased.

'

_'

Diseussion

The

result

shown

in

Figure1(a>

suggests

that

over-lapping

of

the

target

with

the

inducer

in

the

time

domain

is

critical

for

depth

repulsion

but

not neces･ sary

for

depth

attraction.

Although

a simultaneeus

presentation

was

the

best

temporal

condition

for

depth

att'raction,

because

the

depth

induction

index

decreases

as

SOA

increases,

simultaneity

itself

might

not

be

critical

for

depth

attraction,

The

difference

in

temporal

characteristics

between

depth

attraction

and

depth

repulsion

might

refiect

the

possibility

that

'

they

are

mediated

by

different

mechanisms.

'

The

results shown

in

Figure

1

(b)

could

be

interpret-ed as evidence

for

the

importance

of

an

area

of

only

an

inducer

adjacent

to

an area of

both

a

target

and an

inducer.

So

that

this

possibility

could

be

verified a

supplementary

experiment

was

conducted.

The

experirnent used an

inducer

adjacent

to

a

tatget

without an overlapping area.

Depth

repulsion

did

not occur with

this

adjacent

_grating

stimulus.

It

can

be

concluded

that

a

non-overlapping

area

of

inducer,

by

itself,

cloes

not

give

rise

to

depth

repulsion,

Perhaps

a spatio-temporal superimposition

is

rather critical.

There

is

another

possibility.

Due,

to

the

difference

in

shape

between

the

two

_gratings,

the

existence of a non-overlapping

area

might

_prompt

a

target

to

be

perceptually

independent

of

a

superimposed

inclucer

and

so

lead

to

a

transparent

perception.

Given

that

the

depth

perception

of a

target

could

be

infiuenced

by

whether or not

two

_gratings

are

_perceived

as

the

same shape,

it

is

_possible

that

a

relatively

high-]evel

processing,

such

as

shape

perception,

rr}ight

contrib-ute

to

the

induction

of

depth

_perception.

References

'

Smallman,

H.S.

1995

Fine-to-coarse

scale

disam-biguation

in

stereopsis,

Vtsion

ReseaJTh,

35,

1060.Yanagi,

J.,

&

Sato,

T.

I999

Spatial

frequency

tion

in

solving

arnbiguous

binocular

disparity

(in

Japanese

with

English

abstract).

71gze

lkPanese