I was wondering why the function vil_histogram_byte uses a vector of double to populate the histogram information instead of a vector of int. I looked at the source code, and the vector is only incremented by one unit, so it is not possible for the vector to ever hold a floating point number. Why this choice? Is it because double can hold a larger number? Then why not use long (or vxl_int_64)?
Crístian Deives dos Santos Viana [aka CD1]
Google Talk: firstname.lastname@example.org