Ask Your Question

Haarcascade "KILLED"

asked 2013-03-10 08:58:59 -0600

Hello, Im trying to make haar cascade hand detection program,but when i try to train the haarcascade im getting a error of "KILLED".Please find the attached screen shot and suggest me the solution. command given : ./opencv_haartraining -data /home/ubuntu/src/OpenCV*/data/haarcascades -vec samples.vec -bg negatives.dat -nstages 1 -nsplits 1 -minhitrate 0.5 -maxfalsealarm 0.5 -npos 10 -nneg 20 -w 120 -h 120 -nonsym -mem 1000 -mode ALL image description Opencv -2 .4.4 under Ubuntu 12.04 LTS

Thanks in advance

edit retag flag offensive close merge delete

3 answers

Sort by ยป oldest newest most voted

answered 2013-03-11 11:02:57 -0600

A single stage cascaded classifier simply doesn't exist. You can only get a single weak classifier which cannot be a cascaded structure. Knowing this, you should simple select minimum 2 stages for a cascaded classifier.

If you do not understand what i mean, go read the paper of viola&jones, suggesting a cascaded approach for face detection. It will make it more clear for you:

edit flag offensive delete link more

answered 2017-11-06 14:27:30 -0600

classic gravatar image

updated 2017-11-06 14:29:17 -0600

I had the same issue, and it was because I took to much memory for the -precalcValBufSize and -precalcIdxBufSize I think it's same of the "-mem" parameters.

I put for those 2 arguments : -precalcValBufSize 1024 -precalcIdxBufSize 1024

so I reduce it to 256 and then the issue was fixed.

-precalcValBufSize 256 -precalcIdxBufSize 256

edit flag offensive delete link more


This is an interesting read about this topic:

It means Ubuntu thinks the process is taking excessive memory and kills it off. Lowering memory-requirements can indeed prevent this problem.

rhoolwerf gravatar imagerhoolwerf ( 2017-11-06 15:40:37 -0600 )edit

Hmm which system in 2017 still runs on 500MB of memory? o_O

StevenPuttemans gravatar imageStevenPuttemans ( 2017-11-07 02:07:38 -0600 )edit

answered 2013-03-10 12:54:20 -0600

kaipyroami gravatar image

Just an observation, though it might solve the problem. You are using an extremely small nimber of sample images. I am using 2,000+ negatives and 3,000+ positives. Perhaps there are not enough images to start one full stage?

edit flag offensive delete link more

Question Tools


Asked: 2013-03-10 08:58:59 -0600

Seen: 1,707 times

Last updated: Nov 06 '17