# How do i calculate the spread of a cluster of points???

How do I calculate spread each point is from each other.. It has to be added as a feature to tell the program when it has to perform the goodFeatureToTrack Function..

The idea is when the points are too far away from each other I will empty my vector of points And start all over..

Could you show some examples? A first guess could be the variance over the distances from the points to the center of the cluster.

Well.. If You test the lkdemo.cpp within the sample set You Will know the code.. And see how the cluster of points Would look like..