[OpenCV-Python] Change Feature Extraction Method for Object Tracking API

asked 2018-04-15 05:44:01 -0600

I want to change the default observation model (i.e., method of feature extraction) of OpenCV object tracking methods (e.g., BOOSTING, MIL, KCF, TLD, MEDIANFLOW) in python. Would you please kindly help me to do that? (In fact I read the Customizing the CN Tracker but I can't do that in Python)

edit retag flag offensive close merge delete

add a comment

1 answer

Sort by » oldest newest most voted

answered 2018-04-15 05:55:55 -0600

berak
32993 ●7 ●81 ●312

no, you can't do that from python.

you either have to write c++ code, or write your own tracker, e.g. like the MOSSE example here

edit flag offensive delete link

Comments

also note, that the CN Tracker example does not change the observation model, it only adds another feature on top of the existing (CN, gray) ones.

berak ( 2018-04-15 06:06:15 -0600 )edit

add a comment

Links

Question Tools

1 follower

Stats

Asked: 2018-04-15 05:44:01 -0600

Seen: 361 times

Last updated: Apr 15 '18

[OpenCV-Python] Change Feature Extraction Method for Object Tracking API edit