In this e-world the Card (ATM) was becomes the popular way of payment for both online and direct purchase. As the usage increased the fraud rate also increased. To control the fraudulent associated with card we introduce a new method using Hidden Markov Model (HMM). In this we converted the transaction card in series and trained the card initially with the normal usage of the card holder. If it is authentic the transaction process occurs normally to send code to our mobile from that code to credit the amount.If the hacker misuse our card this information were protected by using the code generation to our mobile phone, also can be high-speed processing. However, if codes were generated by unauthorized users, they can trace out some code to a ATM randomly. At that time we propose that the system is appended function for privacy protection to the code use three times chance then it futher wrong code track the Atm card is blocked automatically.The automated
blocking machine is securely accessed using the secure cell phone if the computed transaction code matches the unique security transaction code.