abewley / sort

Simple, online, and realtime tracking of multiple objects in a video sequence.
GNU General Public License v3.0
3.82k stars 1.07k forks source link

I don't understand how states representing the U dot and V dot are updated or predicted? #123

Open Helaly96 opened 3 years ago

Helaly96 commented 3 years ago

Hello, thank you so much for the library. I am implementing a variation of it and including other criteria to the states as well (depth of object )

i have a question tho, for the U dot and V dot , i don't understand how they are predicted. as they are not measured (so obv not corrected in update step) , and neither the State transition matrix nor the measurement matrix updates it by any means.