Hi all. At first sorry for my bad english...
My workaround is windows 7 x86 + opencv 2.4.10
I got 100 positive and 245 negative samples. Then i created s.vec (app shom me message "good.dat(101) : parse errorDone. Created 100 samples" )
opencv_createsamples.exe -info good.dat -vec s.vec -w 20 -h 30
.vec file size is 150kb
then i try to train cascade. i run opencv_traincascade.exe exactly from my haar folder
opencv_traincascade.exe -data haar -vec s.vec -bg bad.dat -numStages 16 -minhitrate 0.999 -maxFalseAlarmRate 0.5 -numPos 100 -numNeg 245 -w 20 -h 30 -precalcValBufSize 2048 -precalcIdxBufSize 2048
and i got error: "Train dataset for temp stage can not be filled. Branch training terminated."
the whole app output is
PARAMETERS: cascadeDirName: haar vecFileName: s.vec bgFileName: bad.dat numPos: 100 numNeg: 245 numStages: 16 precalcValBufSize[Mb] : 2048 precalcIdxBufSize[Mb] : 2048 stageType: BOOST featureType: HAAR sampleWidth: 20 sampleHeight: 30 boostType: GAB minHitRate: 0.995 maxFalseAlarmRate: 0.5 weightTrimRate: 0.95 maxDepth: 1 maxWeakCount: 100 mode: BASIC
===== TRAINING 0-stage ===== POS count : consumed 100 : 100 NEG count : acceptanceRatio 245 : 1 Precalculation time: 4.477 Training until now has taken 0 days 0 hours 0 minutes 6 seconds.
===== TRAINING 1-stage ===== POS count : consumed 100 : 100 NEG count : acceptanceRatio 245 : 0.00395263 Precalculation time: 3.78 Training until now has taken 0 days 0 hours 0 minutes 12 seconds.
===== TRAINING 2-stage ===== POS count : consumed 100 : 100 NEG count : acceptanceRatio 245 : 0.000477984 Precalculation time: 3.784 Training until now has taken 0 days 0 hours 0 minutes 25 seconds.
===== TRAINING 3-stage ===== POS count : consumed 100 : 100 Train dataset for temp stage can not be filled. Branch training terminated.
i read same questions on this forum, but there users got error when training starts. as for me, i see error on 3rd stage (sometimes, with other input dataset on 4 or 5 stages) can somebody tell me what i`m doing wrong?