Sandeep-Sthapit
commented
7 years ago Sorry for the late reply. This was our semester project. Due to limited time we took sample pictures of fist of both hand. As for the palm, we took picture of one hand only (I don't remember if it was left or right) and for the other palm we flipped the picture along vertical axis. This could be the possible reason for the above mentioned problem. Unfortunately, we have not retrained the left hand model. But if you can collect enough samples, you can easily train it yourself.
Hi @Sandeep-777 , Thanks for sharing your code, the project is very useful for me. But I realize the right hand(using "rpalm.xml") detection rate higher than the left hand(using "lpalm.xml"), what is the cause of this problem? Have you retrained the left hand model? Could you please provide the new model of left hand for me if you have retrained it? My email address is ahuljx@126.com Thank you very much! I am looking forward to your reply!
Samylee