hi
I use sdk2.5 usecase --- object detection to do pedestrian detection, and I try to edit its parameters (like block size/ overlap and so on)...
I want to check its result as feature vector ... HOW can I do this way?
ex. on open cv http://lear.inrialpes.fr/people/triggs/pubs/Dalal-cvpr05.pdf
image size(64*128)
block size(16*16)/ cell size(8*8) / block stride(8*8)
then we can get 3780 of features and know their values.
thanks a lot
by Alice