2pm, Tuesday, 31 Jan 2006:

<CENTER><B>
Efficient image approximation with 
geometric dictionaries
</B>

<P>
Rosa Maria Figueras i Ventura
<BR>
LCV/NYU

</CENTER>

Natural images are often modeled through piecewise-smooth regions. 
Region edges, which correspond to the contours of the objects, become, 
in this model, the main information of the signal. Contours have the 
property of being smooth functions along the direction of the edge, and 
irregularities on the perpendicular direction. Modeling edges with the 
minimum possible number of terms is of key importance for numerous 
applications, such as image coding, segmentation or denoising. Standard 
separable basis fail to provide sparse enough representation of 
contours, due to the fact that this kind of basis do not see the 
regularity of edges. This talk shows that the introduction of geometric 
features to the basis functions, such as orientation and arbitrary 
anisotropy, improves the approximation rate. A (highly redundant) 
dictionary having this specific features is then created, and Matching 
Pursuit is used to find a (sparse enough) image approximation in this 
dictionary. The second part of the talk presents a low bit-rate image 
coder based on the previously presented dictionary and Matching Pursuit 
searching algorithm.