SZELISKI BOOK Computer Vision Algorithms Applications Book
<p>Jan 25, 2011 06:10 PM
 
entire book is online and free (this is a 2008 draft;  book published in 2010)
 
Beautiful online book!
 
Fig 1.1  IRIS  flower  gorgeous!
 
page 3  despite advances in CV,  we cannot yet  do what a two year old child can do
(count all the animals in a picture).
 
Why is vision so difficult? In part, it’s because vision is an<b> inverse problem,</b> in which we
seek to <b>recover some unknowns given insufficient information</b> to fully specify the solution. We
must therefore resort to physics-based and probabilistic models to disambiguate between potential
solutions.
 
page 11 history generalized cylinders  Binford;
parts pictorial structures.
 
David Marr 1982  VISION book
 
Computational thy:  GOAL of the the task   (prob.  WHO or <b>WHAT</b> is there; <b>WHERE</b> is it;  Where is it going;
    how far away is it... is it close or far;  is it a threat?
 
Reps and Algos   input, output,  intermediate info, algos.
 
Hardware:  implementation in neurons or transistors.    GPU  and  many core CV systems.
 
 
 
 
 
 
 
 
 
 
 
 
 
 </p>