LTI-Lib (C++ Computer Vision Library) Code

alvarado: Error generating steered channel fixed.

alvarado: Error saving the number of inputs fixed.

alvarado: @FIX: The lti::medianCut functor didn't quantize the colors correctly. They were mapped into a reduced color space, and not mapped back. This problem has now been fixed, but the functor could be revised to provide a much efficient implementation. Any volunteers?

alvarado: @CHG: The serialization of derived classes from lti::thistogram have changed. Now, the written data can also be read as a simple thistogram (1D or 2D), and also a thistogram of 1D or 2D can be read as histogram1D or histogram2D.

wickel: two more features: 1. can autodetect shortest encoding; 2. can write
EPS file to stdout

wickel: fixed errors in comments

wickel: now works with std::ostream, so it can be used with stringstreams

goenner: The tree::node::insertChild(int n) now sets the degree as in the docu.

The other files contain minor changes (see last log-messages) due to cvs-merge.

goenner: Optimization of tline::getIntersectionPointXPol(..).
Fixed tree::node::isChildValid() and iterator tree::getIteratorTo(..).
The later was const!!

Claudia

alvarado: Claudia Goenner added.