Abstract - This paper discusses the method used by facial recognition to access automatic teller machines (ATMs). ATMs are used to perform banking functions such as checking balance, withdrawing money, changing pin numbers, etc. ATM cards and pin numbers are used for security purposes. But this system uses SIM card instead of ATM cards. To increase security, the system first authenticates the person, if it is recognized, it will then ask for the account password. This system used the Spartan 3 FPGA board to control the system. There is a buzzer attached to the FPGA board that instructs the user to log in to the account. If the person is not authenticated, the process is terminated and the output is displayed on the FPGA board with the help of LEDs. Keywords: recognition, ATM, PCA, GSM, FPGA, Euclidean distance I. INTRODUCTION Facial recognition plays a very important role in security system [4]. The main goal of face recognition is to recognize the person from images or videos using face databases. There are many variations to design a face recognition is not an easy task [2]. Due to variations in lighting, facial expression, and poses, facial recognition is difficult to perform [7]. Numerous commercial, defense, and security applications require real-time facial recognition systems, especially when other biometric techniques are not feasible [1]. In this paper, the system uses facial recognition to log in to the ATM. Automated Teller Machines (ATMs) are used by the user to perform banking functions such as withdrawing money, checking balance, etc. Nowadays ATMs are very popular because they operate every day of the week and provide 24-hour service. We can find ATMs everywhere in cities, in train stations, near companies, near restaurants.... .. middle of paper ......otani, Feifei Lee and Tadahiro Ohmi,” Face Recognition Using Self-Organizing Maps” www. intechopen.com[6] Rala M. Ebied, "Feature Extraction using PCA and Kernel-PCA for Face Recognition", in The 8th International Conference on INFOrmatics and Systems Computational Intelligence and Multimedia Computing Track, 2012, pp mm72-mm77[7] Mohammed Alwakeel, Zyad Shaaban, "Face recognition based on Haar wavelet transform and principal component analysis via Levenberg-Marquardt backpropagation neural network" in European Journal of Scientific Research, 2010 pp. 25-31[8] Kyungnam Kim, “Face Recognition using Principle Component”, Analysis” www.google.com[9] Kyungnam Kim, “Face Recognition using Principle Component”, Analysis” www.google.com[10] Lindsay I Smith, “A tutorial on Principal Components Analysis” February 26, 2002 [11] www.mathwork.com[12] www.google.com