Image translation and rotation reduces the accuracy of hand vein recognition. Aiming at this problem, a new hand vein recognition algorithm was proposed based on multi-feature fusion. The characteristic of the approach was to combine local and global features for hand vein recognition. Firstly, intersection points and endpoints were selected as feature points. The reference point for image matching was extracted from feature points. The relative distances between the reference points to feature points were computed. The angles between the adjacent connections were calculated and used as local features. Then the moment invariants were calculated as global features. Finally these features were combined for hand vein recognition. Experimental results show that the proposed algorithm is able to achieve hand vein recognition reliably and quickly.