Endless loop: Haar, LBP, HOG traincascade opencv stuck

Question

I am trying to create a classifier for detecting faces in thermal imaging images. Therefore, I tried to train using the Haar, LBP, and HOG classifiers. I work with OpenCV 2.4.8 on windows.

opencv_traincascade.exe -data haarcascades -vec pos.vec -bg neg.txt -numPos 250 -numStages 24 -numNeg 900 -w 24 -h 24

I have 307 positive samples. Negative samples have a size of 75x75. For each of the three cases, the training gets stuck at a certain stage - earlier for Haar (stage 12), and then for LBP (stage-14/15). I reduced the number of negatives (to 200), but that means the training stuck at a later stage. Training has not progressed from 2 days. No negatives are consumed, and the command window looks like this:

===== TRAINING 14-stage =====
<BEGIN
POS count : consumed   255 : 262

Also

• What does the POS counter consume and consume the amount of NEG?
• When I decrease minHitRate to say 0.7, why does the amount of POS consumed increase?

Please let me know what I am doing wrong. Thank.

Answer 1

. , , . , , , , . , , , - . , .

Answer 2

- , . 1000 2000 ,