opencv - TrainCascade stuck on getting new negatives -


i'm working opencv 2.4.7 on windows. i'm using traincascade train new haar cascade eyeglasses using following command:

opencv_traincascade -data traincascade20 -vec vector3.vec -bg infofile3.txt -numstages 40 -minhitrate 0.999  maxfalsealarmrate 0.5 -numpos 170 -numneg 1000 -w 20 -h 20 -mode -precalcvalbufsize 1024 -precalcidxbufsize 1024 

it's stuck (or progressing slow) on stage 24 on phase of getting new negatives. negative images file "infofile3.txt" contains 12k negative image.

can please explain why it's progressing , can make progress (a lot) faster?

thanks in advance,

gil.

around 24 hours sounds normal me. haar training can take days depending on size , number of samples. , of course on computer well. longest training took approximately week hand detection.

if worried, check whether haar training still on-going, can try generate intermediate haar cascade xml file, data available. if able generate xml file, show it's still running(albeit slow) , not stuck.

how improve haar training speed, solution know or used before "paralleling"

a quick search on google leads few link, here's 1 of them: http://www.computer-vision-software.com/blog/2009/06/parallel-world-of-opencv/

i have used such methods, , it's pretty efficient in cutting time taken train haar cascade. hope method suits well. try method of generating immediate xml file current data available first though. if there needs, comment, try soon. cheers.


Comments

Popular posts from this blog

html - Sizing a high-res image (~8MB) to display entirely in a small div (circular, diameter 100px) -

java - IntelliJ - No such instance method -

identifier - Is it possible for an html5 document to have two ids? -