The Real-time Attendance Maintaining System is difficult process if it is done manually. It is an important requirement in the modern era to improve the working efficiency of the entire private and public sector. There are many smart and automated forms of biometrics to monitor the movement of students or employees such as fingerprint, palm scanning, iris,voice recognition etc. Face Recognition is one of them. By using this system, the issue of fake attendance and proxies can be solved. This paper suggests a robust procedure for monitoring attendees using facial recognition. The method proposed here for real-time character identification involves a facial recognition technique using Haar-like features with a Cascade classifier. For the purpose of face recognition histogram of local binary patterns (LBPH) are used. The major steps in the proposed system are capturing the images of the students, storing the captured images in the database, detecting the faces and recognizing them. The comparison is done by crosschecking the detected face with the database of student’s faces. This smart system will be effective way to maintain the attendance and records of students.