I am using opencv_traincascade for object detection. I try to find glasses on photo. For this, I've downloaded 830 pictures like this: http://pi1.lmcdn.ru/product/V/I/VI060DWIHZ27_1_v2.jpg
Then I've downloaded many pictures with model in dresses or just dresses photos, 1799 photos.
Then I've start opencv_traincascade with parameters: opencv_traincascade -data Feature/classifier -vec samples.vec -bg negatives.txt -numStages 10 -minHitRate 0.999 -maxFalseAlarmRate 0.5 -numPos 830 -numNeg 1799 -w 60 -h 90 -mode ALL -precalcValBufSize 1024 -precalcIdxBufSize 1024
But after step 4, I have a message: Train dataset for temp stage can not be filled. Branch training terminated.
The full stacktrace is:
➜ pictureFeature opencv_traincascade -data Feature/classifier -vec samples.vec -bg negatives.txt -numStages 10 -minHitRate 0.999 -maxFalseAlarmRate 0.5 -numPos 830 -numNeg 1799 -w 60 -h 90 -mode ALL -precalcValBufSize 1024 -precalcIdxBufSize 1024 PARAMETERS: cascadeDirName: Feature/classifier vecFileName: samples.vec bgFileName: negatives.txt numPos: 830 numNeg: 1799 numStages: 10 precalcValBufSize[Mb] : 1024 precalcIdxBufSize[Mb] : 1024 acceptanceRatioBreakValue : -1 stageType: BOOST featureType: HAAR sampleWidth: 60 sampleHeight: 90 boostType: GAB minHitRate: 0.999 maxFalseAlarmRate: 0.5 weightTrimRate: 0.95 maxDepth: 1 maxWeakCount: 100 mode: ALL ===== TRAINING 0-stage ===== <BEGIN POS count : consumed 830 : 830 NEG count : acceptanceRatio 1799 : 1 Precalculation time: 26 +----+---------+---------+ | N | HR | FA | +----+---------+---------+ | 1| 1| 1| +----+---------+---------+ | 2| 1| 1| +----+---------+---------+ | 3| 1| 0.145636| +----+---------+---------+ END> Training until now has taken 0 days 5 hours 22 minutes 10 seconds. ===== TRAINING 1-stage ===== <BEGIN POS count : consumed 830 : 830 NEG count : acceptanceRatio 1799 : 0.145715 Precalculation time: 24 +----+---------+---------+ | N | HR | FA | +----+---------+---------+ | 1| 1| 1| +----+---------+---------+ | 2| 1| 1| +----+---------+---------+ | 3| 1| 1| +----+---------+---------+ | 4| 1| 0.762646| +----+---------+---------+ | 5| 1| 0.432462| +----+---------+---------+ END> Training until now has taken 0 days 14 hours 38 minutes 28 seconds. ===== TRAINING 2-stage ===== <BEGIN POS count : consumed 830 : 830 NEG count : acceptanceRatio 1799 : 0.062696 Precalculation time: 28 +----+---------+---------+ | N | HR | FA | +----+---------+---------+ | 1| 1| 1| +----+---------+---------+ | 2| 1| 1| +----+---------+---------+ | 3| 1| 1| +----+---------+---------+ | 4| 1| 0.590328| +----+---------+---------+ | 5| 1| 0.187326| +----+---------+---------+ END> Training until now has taken 0 days 23 hours 21 minutes 4 seconds. ===== TRAINING 3-stage ===== <BEGIN POS count : consumed 830 : 830 NEG count : acceptanceRatio 1799 : 0.0117929 Precalculation time: 21 +----+---------+---------+ | N | HR | FA | +----+---------+---------+ | 1| 1| 1| +----+---------+---------+ | 2| 1| 1| +----+---------+---------+ | 3| 1|0.0944969| +----+---------+---------+ END> Training until now has taken 1 days 3 hours 47 minutes 34 seconds. ===== TRAINING 4-stage ===== <BEGIN POS count : consumed 830 : 830 NEG count : acceptanceRatio 1799 : 0.00112161 Precalculation time: 18 +----+---------+---------+ | N | HR | FA | +----+---------+---------+ | 1| 1| 0| +----+---------+---------+ END> Training until now has taken 1 days 5 hours 4 minutes 35 seconds. ===== TRAINING 5-stage ===== <BEGIN POS count : consumed 830 : 830 Train dataset for temp stage can not be filled. Branch training terminated. I tried to use cascade.xml for object searching, but the result was completely fail.
Could somebody help with my problem?
2 Answers
Answers 1
If you look at the error you'll find that it stopped at 'NEG count' which means there was an issue with reading the negative training data set images at stage-5. So you need to fix the path to the negative training samples for it to work.
Here is a detailed discussion on this issue which might be helpful: Cascade Training Error OpenCV 2.4.4
It can also be an issue with the bg.txt file, here is another answer: error in train casacde.
Another post with similar error: Train dataset for temp stage can not be filled.
If nothing works try using a different version of open cv: http://sourceforge.net/projects/opencvlibrary/files/opencv-unix/2.4.9/opencv-2.4.9.zip/download
Answers 2
That's a pretty common error. There is not much I know about this error and I can only guess a solution, since there are a lot of possible reasons for that error.
It could be relative to the directories structure (just notice that the path for negative examples should be relative to the current directory)
or because the line break in bg file (negatives.txt in your case) are bad encoded. Could you please double check that last character of each line in the 'negatives.txt' file are '\n' and not '\r' ?
Also keep trying to maintain positive to negative samples ratio around 25%-30%
And finaly be sure that all your positive and negative images contain the head face:
positive = headface with sunglasses
negative = headface only
Regards
0 comments:
Post a Comment