Vision-based precise cash counting in ATM machines

The demand to use ATM machines has increased tremendously worldwide during the last decade. Traditional cash counting in ATM machines is based on the mechanical mechanism that is in touch with cash and may cause some damages from time to time. This paper proposes a novel visual based counting technology for cash counting in ATM machines. To achieve precise counting, intelligent decision making algorithms are deployed to evaluate the counting performance so that the machine will not produce incorrect counts. The experimental results showed that the proposed method can eliminate false positive counts effectively.