This paper proposes a novel contactless biometrie system for multisampling hand recognition. We include a novel acquisition device for contactless hand recognition and a study about the multi-sample acquisition as a way to improve the performance. Popular features extraction methods for hand geometry and palmprint are studied and a database of 100 people with more than 2000 hand is employed for the experimentation. The results suggest how a multi-sample approach outperforms the traditional single sample approach with improvements around 47% and EER of 0.21%.