Sample size
Whether there will be accepted a recognition in case of the identical sizes at positive and negative examples?
Whether there will be accepted a recognition in case of the identical sizes at positive and negative examples?
Asked: 2014-03-07 09:42:10 -0600
Seen: 138 times
Last updated: Mar 07 '14
can you switch out xml files facedetection object recognition
Verify vector file and background file
acceptanceRatio different when restarting traincascade
OpenCV Error : Assertion failed during haartraining
Problems with Traincascade: A Practical Problem
About image backgrounds while preparing training dataset for cascaded classifier [closed]
negative images in haarcascade - training
traincascade - "Cascade classifier can't be trained. Check the used training parameters."
It is always advisable to keep the sizes of positive and negative images same and small.Width can range from 20pixels to 100pixels and same goes with the height.The image must contain only the required important stuff.