33 |
\begin{document} |
\begin{document} |
34 |
\maketitle |
\maketitle |
35 |
|
|
36 |
\font\foofonti = cmfrak scaled 2000 |
%\font\foofonti = cmfrak scaled 2000 |
37 |
\font\foofontii = qhvb scaled 2000 |
%\font\foofontii = qhvb scaled 2000 |
38 |
|
|
39 |
|
|
40 |
\begin{abstract} |
\begin{abstract} |
264 |
perceptually, for visualizing surface orientation\cite{schweitzer83texturing,interrante97illustrating} and scalar or vector fields\cite{ware95texture}, |
perceptually, for visualizing surface orientation\cite{schweitzer83texturing,interrante97illustrating} and scalar or vector fields\cite{ware95texture}, |
265 |
and statistically, as samples from a probability distribution on a random field |
and statistically, as samples from a probability distribution on a random field |
266 |
\cite{cross83markov,geman84stochastic}. |
\cite{cross83markov,geman84stochastic}. |
267 |
|
% XXX: there's overlap between the enumerated cases |
268 |
|
|
269 |
%Textures have also been modeled statistically, |
%Textures have also been modeled statistically, |
270 |
%as samples from a probability distribution on a random field. |
%as samples from a probability distribution on a random field. |
274 |
%depends only on the values of its neighborhood (local characteristics). |
%depends only on the values of its neighborhood (local characteristics). |
275 |
%XXX: resolution-dependency? |
%XXX: resolution-dependency? |
276 |
|
|
277 |
% In this article, we apply texture shading to synthesize a large number |
%% XXX: this is not really texturing: |
|
% of unique textures for distinguishing virtual objects. |
|
|
|
|
|
\subsection{Texture perception} |
|
|
|
|
|
Psychophysical studies on texture perception have mostly concentrated |
|
|
on pre-attentive |
|
|
\emph{visual texture discrimination}\cite{julesz62visualpattern}, |
|
|
the ability of human observers to effortlessly discriminate |
|
|
pairs of certain textures (see Bergen\cite{bergen91theories} for a review). |
|
|
%The term is often used interchangably with \emph{texture segregation}, |
|
|
%the more specific task of finding the border between differently textured |
|
|
%areas (different phases of local characteristics at the |
|
|
%border can segregate otherwise indiscriminable textures). |
|
|
% |
|
|
%First experiments on computer-generated, unnatural textures in the 60s |
|
|
%\cite{julesz62visualpattern} led to proposals of discrimination models |
|
|
%based on the $N$th-order statistics of textures |
|
|
%(the joint distributions of the values at the corners of a randomly |
|
|
%placed (translated) $N$-gon for all different $N$-gons). |
|
|
%%and connectivity structures of certain micropatterns. |
|
|
% |
|
|
First discrimination models were based |
|
|
on the $N$th-order statistics of textures |
|
|
(the joint distributions of the values at the corners of a randomly |
|
|
placed (translated) $N$-gon for all different $N$-gons). |
|
|
However, the order of similarity in the statistics did not |
|
|
consistently explain discrimination performance, and certain |
|
|
pre-attentive local features were conjectured. |
|
|
|
|
|
Julesz\cite{julesz81textons} proposed that texture discrimination could be |
|
|
explained by the densities of textons, fundamental texture elements, such as |
|
|
elongated blobs, line terminators, and line crossings. |
|
|
However, the textons are hard to define formally. |
|
|
|
|
|
Much simpler filtering-based models can explain texture discrimination |
|
|
just as well \cite{bergen88earlyvision}. |
|
|
In these models, a bank of linear filters is applied to the texture followed |
|
|
by a nonlinearity and then another set of filters to extract features |
|
|
(see, e.g., \cite{heeger95pyramid} for an application). |
|
|
%In \cite{heeger95pyramid}, new textures with appearance similar |
|
|
%to a given texture are created by matching certain histograms |
|
|
%of filter responses. |
|
|
|
|
|
XXX: higher-level pre-attentive processes? |
|
|
|
|
|
%XXX: texture perception reviews |
|
|
|
|
278 |
There have been studies on |
There have been studies on |
279 |
mapping texture appearance to an Euclidian texture space |
mapping texture appearance to an Euclidian texture space |
280 |
(see \cite{gurnsey01texturespace} and the references therein): |
(see \cite{gurnsey01texturespace} and the references therein): |
288 |
sufficient \cite{rao96texturenaming}, but often semantic connections cause the |
sufficient \cite{rao96texturenaming}, but often semantic connections cause the |
289 |
similarity to be context-dependant, making it hard to assess the |
similarity to be context-dependant, making it hard to assess the |
290 |
dimensionality. |
dimensionality. |
291 |
% XXX: this is something we should experiment with our textures |
%% XXX: this is something we should experiment with our textures |
292 |
|
|
293 |
|
% In this article, we apply texture shading to synthesize a large number |
294 |
|
% of unique textures for distinguishing virtual objects. |
295 |
|
|
296 |
|
%\subsection{Texture perception} |
297 |
|
% |
298 |
|
%Psychophysical studies on texture perception have mostly concentrated |
299 |
|
%on pre-attentive |
300 |
|
%\emph{visual texture discrimination}\cite{julesz62visualpattern}, |
301 |
|
%the ability of human observers to effortlessly discriminate |
302 |
|
%pairs of certain textures (see Bergen\cite{bergen91theories} for a review). |
303 |
|
%%The term is often used interchangably with \emph{texture segregation}, |
304 |
|
%%the more specific task of finding the border between differently textured |
305 |
|
%%areas (different phases of local characteristics at the |
306 |
|
%%border can segregate otherwise indiscriminable textures). |
307 |
|
%% |
308 |
|
%%First experiments on computer-generated, unnatural textures in the 60s |
309 |
|
%%\cite{julesz62visualpattern} led to proposals of discrimination models |
310 |
|
%%based on the $N$th-order statistics of textures |
311 |
|
%%(the joint distributions of the values at the corners of a randomly |
312 |
|
%%placed (translated) $N$-gon for all different $N$-gons). |
313 |
|
%%%and connectivity structures of certain micropatterns. |
314 |
|
%% |
315 |
|
%First discrimination models were based |
316 |
|
%on the $N$th-order statistics of textures |
317 |
|
%(the joint distributions of the values at the corners of a randomly |
318 |
|
%placed (translated) $N$-gon for all different $N$-gons). |
319 |
|
%However, the order of similarity in the statistics did not |
320 |
|
%consistently explain discrimination performance, and certain |
321 |
|
%pre-attentive local features were conjectured. |
322 |
|
% |
323 |
|
%Julesz\cite{julesz81textons} proposed that texture discrimination could be |
324 |
|
%explained by the densities of textons, fundamental texture elements, such as |
325 |
|
%elongated blobs, line terminators, and line crossings. |
326 |
|
%However, the textons are hard to define formally. |
327 |
|
% |
328 |
|
%Much simpler filtering-based models can explain texture discrimination |
329 |
|
%just as well \cite{bergen88earlyvision}. |
330 |
|
%In these models, a bank of linear filters is applied to the texture followed |
331 |
|
%by a nonlinearity and then another set of filters to extract features |
332 |
|
%(see, e.g., \cite{heeger95pyramid} for an application). |
333 |
|
%%In \cite{heeger95pyramid}, new textures with appearance similar |
334 |
|
%%to a given texture are created by matching certain histograms |
335 |
|
%%of filter responses. |
336 |
|
% |
337 |
|
%XXX: higher-level pre-attentive processes? |
338 |
|
% |
339 |
|
%%XXX: texture perception reviews |
340 |
|
% |
341 |
|
%There have been studies on |
342 |
|
%mapping texture appearance to an Euclidian texture space |
343 |
|
%(see \cite{gurnsey01texturespace} and the references therein): |
344 |
|
%in the reported experiments, three dimensions have been sufficient |
345 |
|
%to explain most of the variation in the similarity judgements for |
346 |
|
%artificial textures. |
347 |
|
%However, the texture stimuli have been somewhat simple |
348 |
|
%(no color, lack of frequency-band interaction, etc.). |
349 |
|
%For some natural texture sets, |
350 |
|
%three dimensions have also been |
351 |
|
%sufficient \cite{rao96texturenaming}, but often semantic connections cause the |
352 |
|
%similarity to be context-dependant, making it hard to assess the |
353 |
|
%dimensionality. |
354 |
|
%% XXX: this is something we should experiment with our textures |
355 |
|
|
356 |
\subsection{Focus+Context views} |
\subsection{Focus+Context views} |
357 |
|
|
542 |
we have to take into account the properties of the human |
we have to take into account the properties of the human |
543 |
visual system. |
visual system. |
544 |
|
|
545 |
% The seed for randomly choosing |
Psychophysical studies on texture perception have mostly concentrated |
546 |
% an easily distinguishable unique background from a |
on pre-attentive |
547 |
% distribution based on |
\emph{visual texture discrimination}\cite{julesz62visualpattern}, |
548 |
|
the ability of human observers to effortlessly discriminate |
549 |
|
pairs of certain textures (see Bergen\cite{bergen91theories} for a review). |
550 |
|
Nevertheless, |
551 |
|
discrimination models can provide insight on the pre-attentive |
552 |
|
processes underlying global perception. |
553 |
|
|
554 |
%providing an infinite source of unique backgrounds. |
Julesz\cite{julesz81textons} proposed that texture discrimination could be |
555 |
%generating textures based on seed numbers [identity] |
explained by the densities of textons, fundamental texture elements, such as |
556 |
The first stages |
elongated blobs, line terminators, and line crossings. |
557 |
of visual perception |
However, the textons are hard to define formally. |
558 |
are fairly well known |
|
559 |
(see, e.g.,~Bruce et al\cite{bruce96visualperception}): |
Much simpler filtering-based models can explain texture discrimination |
560 |
|
just as well \cite{bergen88earlyvision}. |
561 |
|
In these models, a bank of linear filters is applied to the texture followed |
562 |
|
by a nonlinearity and then another set of filters to extract features |
563 |
|
(see, e.g., Heeger\cite{heeger95pyramid}). |
564 |
|
There is also physiological evidence of filtering processes: |
565 |
|
%The first stages |
566 |
|
%of visual perception |
567 |
|
%are fairly well known |
568 |
in the visual cortex, there are cells sensitive to different |
in the visual cortex, there are cells sensitive to different |
569 |
frequencies, orientations, and locations in the visual field. |
frequencies, orientations, and locations in the visual field |
570 |
|
(see, e.g.,~Bruce et al\cite{bruce96visualperception}). |
571 |
|
|
572 |
On a higher level, the correlations between local features are combined |
On a higher level, the correlations between local features are combined |
573 |
by forming contours and possibly |
by forming contours and possibly |
574 |
other higher-level constructions. |
other higher-level constructions (see, e.g., \cite{saarinen97integration}). |
575 |
These higher levels are not yet thoroughly understood; |
These higher levels are not yet thoroughly understood; |
576 |
some theories |
some theories |
577 |
(see, e.g., Biederman\cite{biederman87}) |
(see, e.g., Biederman\cite{biederman87}) |
578 |
assume certain primitive shapes whose |
assume certain primitive shapes whose |
579 |
structure facilitates recognition. |
structure facilitates recognition. |
580 |
|
|
581 |
|
% The seed for randomly choosing |
582 |
|
% an easily distinguishable unique background from a |
583 |
|
% distribution based on |
584 |
|
|
585 |
|
%providing an infinite source of unique backgrounds. |
586 |
|
%generating textures based on seed numbers [identity] |
587 |
|
|
588 |
\begin{figure} |
\begin{figure} |
589 |
\centering |
\centering |
590 |
%\fbox{\vbox{\vskip 3in}} |
%\fbox{\vbox{\vskip 3in}} |