/[gzz]/manuscripts/Paper/paper.tex
ViewVC logotype

Diff of /manuscripts/Paper/paper.tex

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 1.19 by jvk, Wed Mar 12 09:16:28 2003 UTC revision 1.20 by jvk, Wed Mar 12 15:25:15 2003 UTC
# Line 251  Psychological studies on texture percept Line 251  Psychological studies on texture percept
251  on \emph{texture discrimination}, the ability of human observers to  on \emph{texture discrimination}, the ability of human observers to
252  discriminate pairs of textures.    discriminate pairs of textures.  
253  The term is often used interchangably with \emph{texture segregation},  The term is often used interchangably with \emph{texture segregation},
254  the more specific task of finding the border between areas of  the more specific task of finding the border between differently textured
255  different textures (different phases of local characteristics at the  areas (different phases of local characteristics at the
256  border can segregate otherwise indiscriminable textures).  border can segregate otherwise indiscriminable textures).
257    
258  First experiments on computer-generated, unnatural textures in the 60s  First experiments on computer-generated, unnatural textures in the 60s
259  \cite{julesz62visualpattern} led to proposals of discrimination models  \cite{julesz62visualpattern} led to proposals of discrimination models
260  based on $N$th-order statistics (the joint distributions of all  based on the $N$th-order statistics of textures
261  $N$-tuples of pixels for given $N$) and connectivity structures of  (the joint distributions of the values at the corners of a randomly
262  certain micropatterns.  placed (translated) $N$-gon for all different $N$-gons).
263    %and connectivity structures of certain micropatterns.
264    
265  Statistical modeling of textures as samples from a probability  Statistical modeling of textures as samples from a probability
266  distribution on a random field as already seen in \cite{julesz62visualpattern}  distribution on a random field as already seen in \cite{julesz62visualpattern}
# Line 269  The most popualar computational approach Line 270  The most popualar computational approach
270  depends only on the values of its neighborhood (local characteristics).  depends only on the values of its neighborhood (local characteristics).
271  XXX: resolution-dependency?  XXX: resolution-dependency?
272    
273  Attempt to explain texture perception by the densities of textons  Attempt to explain texture discrimination by the densities of textons
274  \cite{julesz81textons}, fundamental texture elements, such as  \cite{julesz81textons}, fundamental texture elements, such as
275  elongated blobs, line terminators, line crossings, etc.    elongated blobs, line terminators, line crossings, etc.  
276  However, the textons are hard to define formally.  However, the textons are hard to define formally.
277    
278  Of course, such models are not directly applicable on high-resolution  Much simpler filtering-based models can explain texture discrimination
279  textures; some kind of filtering would be required to obtain the input.    just as well \cite{bergen88earlyvision}.
 However, the filtering itself can also have good explanatory power  
 \cite{bergen88earlyvision}.XXX  
   
 Filtering based approach, e.g., \cite{heeger95pyramid}.  
280  Essentially a bank of linear filters is applied to the texture followed  Essentially a bank of linear filters is applied to the texture followed
281  by a nonlinearity and then another set of filters.  by a nonlinearity and then another set of filters.
282    In \cite{heeger95pyramid}, new textures with appearance similar
283    to a given texture are created by matching certain histograms
284    of filter responses.
285    
286  Mapping texture appearance to an Euclidian texture space  Mapping texture appearance to an Euclidian texture space
287  (see \cite{gurnsey01texturespace} and the references therein):  (see \cite{gurnsey01texturespace} and the references therein):
# Line 292  However, the texture stimuli have been s Line 292  However, the texture stimuli have been s
292  (no color, lack of frequency-band interaction, etc.).  (no color, lack of frequency-band interaction, etc.).
293  For some natural texture sets (see, e.g., \cite{rao96texturenaming}),  For some natural texture sets (see, e.g., \cite{rao96texturenaming}),
294  three dimensions have also been  three dimensions have also been
295  sufficient, but often the semantic connections cause the  sufficient, but often semantic connections cause the
296  similarity to be context-dependant, making it hard to assess the  similarity to be context-dependant, making it hard to assess the
297  dimensionality.  dimensionality.
298  % XXX: this is something we should experiment with our textures  % XXX: this is something we should experiment with our textures
# Line 301  XXX: reviews Line 301  XXX: reviews
301    
302  XXX: physiological knowledge of visual perception  XXX: physiological knowledge of visual perception
303    
304    XXX: in most work, texture is considered as the output of a stochastic
305    process that produces certain repeating features.
306    Different samples from the process are considered as the same texture.
307    The textures created by our algorithm, although repeating, are more like
308    complete images rather than microstructure.
309    Therefore, higher level processes of vision are also involved
310    in the perception and recognition.
311    
312  \subsection{Focus+Context views}  \subsection{Focus+Context views}
313    
314  Focus+Context, or, fisheye views\cite{fc-fisheye} are  Focus+Context, or, fisheye views\cite{fc-fisheye} are

Legend:
Removed from v.1.19  
changed lines
  Added in v.1.20

savannah-hackers-public@gnu.org
ViewVC Help
Powered by ViewVC 1.1.26